СИСТЕМЫ И СПОСОБЫ ДЛЯ СОЗДАНИЯ ТРЕХМЕРНОГО ТЕКСТУРНОГО АТЛАСА

Изобретение относится, в общем, к системам и способам для создания трехмерного (3D) текстурного атласа. Техническим результатом является повышение эффективности использования кэш-памяти текстур посредством уменьшения объема кэш-памяти текстур, необходимого для хранения каждого текстурного атласа. В способе создания трехмерного текстурного атласа подразделяют трехмерную треугольную сетку на множество скорректированных треугольников. Группируют множество скорректированных треугольников в группы скорректированных треугольников. Сканируют каждую группу скорректированных треугольников внутри трехмерного ограничивающего объема, чтобы сформировать стандартизованную текстурную мозаику или нестандартизованную текстурную мозаику. Уменьшают объем компьютерной кэш-памяти текстур, требуемой для хранения текстурного атласа, путем объединения каждой стандартизованной текстурной мозаики, чтобы сформировать трехмерный текстурный атлас, содержащий равное число стандартизованных текстурных мозаик в каждом ...

ИНТЕРФЕЙСЫ ВИЗУАЛЬНОГО ОБЪЕКТА И ГРАФА СЦЕНЫ

Изобретение относится к компьютерным системам, в частности к обработке графической и другой видеоинформации для отображения на компьютерных системах. Техническим результатом изобретения является улучшение технологии обработки графики. Этот результат достигается тем, что способ, реализуемый в вычислительной среде, в которой принимается программный код для одного или более объектов, и предназначенный для обеспечения сопряжения со структурой данных графа сцены для создания графики, содержит этапы, на которых принимают вызов функции, соответствующий программному коду на высокоуровневом языке разметки графики, на уровне визуального интерфейса прикладного программирования в формате, который обеспечивает возможность непосредственного размещения этого функционального вызова на упомянутый уровень интерфейса прикладного программирования; отвечают на вызов функции посредством модифицирования данных в графе сцены, активируют средство управления визуальными объектами для обхода графа сцены и визуализации ...

СПОСОБ И СИСТЕМА ДЛЯ ПРЕДСКАЗАНИЯ БУДУЩЕГО СОБЫТИЯ В БЕСПИЛОТНОМ АВТОМОБИЛЕ (SDC)

Dreidimensionale Beschriftungen für Strassenansichtsdaten

System zum Erzeugen und Anzeigen von Beschriftungen, umfassend: einen Server, der einen Beschriftungsgenerator beinhaltet, worin der Beschriftungsgenerator eine erste Beschriftung erzeugt durch das Projizieren eines Auswahlobjekts, das bei Ansehen eines ersten Bildes von einem Benutzer eingegeben wird, auf ein dreidimensionales Modell, das mehreren zweidimensionalen Bildern entspricht, das Bestimmen eines Standorts, der der Projektion auf dem dreidimensionalen Modell entspricht, und das Assoziieren des Standorts mit Inhalt, der von einem Benutzer eingegeben wird; und eine Beschriftungsdatenbank, die mit dem Server verbunden ist, die Beschriftungen, die durch den Beschriftungsgenerator erzeugt wurden, speichert, worin der Server die erste Beschriftung von der Beschriftungsdatenbank abruft und die erste Beschriftung an einen Browser zur Anzeige weiterleitet.

VORRICHTUNG UND VERFAHREN FÜR ASYNCHRONES RAYTRACING

Vorrichtung und Verfahren für asynchrone Raytracing. Zum Beispiel weist eine Ausführungsform eines Prozessors auf: einen Begrenzungsvolumenhierarchie-(BVH) Generator zum Konstruieren einer BVH, aufweisend mehrere hierarchisch angeordnete Knoten, aufweisend einen Wurzelknoten, mehrere interne Knoten und mehrere Blattknoten, aufweisend Primitive, wobei jeder interne Knoten einen Kindknoten an entweder dem Wurzelknoten oder einem anderen internen Knoten aufweist und jeder Blattknoten einen Kindknoten an einem internen Knoten aufweist; eine erste Datenspeicherbank, die als eine erste Vielzahl von Einträgen anzuordnen ist; eine zweite Datenspeicherbank, die als eine zweite Vielzahl von Einträgen anzuordnen ist, wobei jeder Eintrag der ersten Vielzahl von Einträgen und der zweiten Vielzahl von Einträgen einen Strahl speichern soll, der die BVH durchqueren soll; eine Zuordnungsschaltung zum Verteilen eines eingehenden Strahls zu entweder der ersten Datenspeicherbank oder der zweiten Datenspeicherbank ...

Objektverfolgung

Verfahren und Systeme werden zum Verfolgen eines Objekts bereitgestellt. Das System beinhaltet ein Datenempfangsmodul, das dazu konfiguriert ist, dreidimensionale Bildgebungsdaten und zweidimensionale Bildgebungsdaten zu empfangen. Ein dreidimensionales Objektidentifikationsmodul ist konfiguriert, ein dreidimensionales Objekt abzugrenzen und dreidimensionale Objektdaten auszugeben. Ein Einstellmodul ist dazu konfiguriert, die dreidimensionalen Objektdaten basierend auf den zweidimensionalen Objektdaten einzustellen und eingestellte dreidimensionale Objektdaten auszugeben. Ein Verfolgungsmodul ist dazu konfiguriert, mindestens ein Objekt unter Verwendung der eingestellten dreidimensionalen Objektdaten zu verfolgen.

Hybrid hierarchy for ray tracing

Clipping of graphics primitives

Conservative Rasterization Using Gradients

Reorientation of cardiac PET images

Graphics processing systems

Techniques for interior coordinates

Graphics processing systems

3D scene rendering

Tiling a primitive in a graphics processing system

Tiling a primitive 904 in a graphics processing system to determine which of a plurality of tiles 902 of a rendering space the primitive (e.g. a triangle or polygon) intersects or overlaps with. The tiling involves testing tiles in a subset of tiles to determine whether they overlap, cover or intersect the primitive. Tiles in regions bounded by tested tiles are determined to overlap with primitive based on the test results of the tested tiles. Preferably the test may involve using an edge equation to determine whether certain points of the tile are inside or outside the primitive. A bounding box around the primitive may also be determined such that tiling tests are only conducted for tiles intersecting the bounding box. Preferably the subset of tiles to be tested comprises tiles spaced at regular intervals (more preferably one tested tile in each 2-by-2 block of tiles) and may contain the four corner tiles of the bounding box. The tests may be performed on a sparse subset, and then optionally ...

TRANSPORTER SEGMENTATION FOR DATA CAPTURE SYSTEM

Performing non-maximum suppression in parallel

The application describes apparatuses, systems, and techniques to perform non-maximum suppression (NMS) in parallel to remove redundant bounding boxes 606. In at least one embodiment, two or more parallel circuits to perform two or more portions of an NMS algorithm in parallel to remove one or more redundant bounding boxes 606 corresponding to one or more objects within one or more digital images 601.

Intersection testing in a ray tracing system using packets comprising indications of rays and/or bundles of rays

Path Optimisation

OPERATOR ASSISTANCE VISION SYSTEM

A vision system (200) for assisting an operator of a machine (100) is provided. An image capturing device (120), mounted on the machine (100), is configured to capture an image (302) of an environment of the machine (100). The vision system (200) includes a controller (202) configured to apply an object detection process to detect an object (122) in the image (302). The controller (202) determines a bounding box (402) comprising one or more pixels associated with the object (122). The controller (202) determines a height and a range associated with the object (122) based on the bounding box (402). The controller (202) extracts the one or more pixels within the bounding box (402). The controller (202) is further configured to generate a three-dimensional (3D) view (500, 600) comprising the object (122) based on the image captured by the image capturing device (120). The controller (202) reinserts the one or more pixels as a vertical pop-up element with respect to a ground plane in the 3D ...

SYSTEMS AND METHODS FOR GENERATING PROJECTION IMAGES

Systems, methods, and non-transitory computer-readable media taught herein process a subset of a three-dimensional array of voxels to form an image of a region of interest in an object. Images thus formed are clear, comprehensible, and contextual. The systems, methods, and non-transitory computer-readable media taught herein select the subset of voxels that represent the region of interest along a direction perpendicular to a view direction defined by the region of interest. The subset of voxels represents one or more portions of a plurality of image slices of the object.

PLACEMENT OF AUGMENTED REALITY OBJECTS USING A GUIDE MARKER

In some implementations, a system may include a camera, a display, one or more memories, and one or more processors communicatively coupled to the one or more memories. The system may output a guide marker for presentation on the display. The guide marker may be superimposed on an image being captured by the camera and presented on the display. The system may determine three-dimensional (3D) coordinates of an intersection point associated with the guide marker. The intersection point may be a point where a projection of the guide marker into 3D space intersects with a horizontal plane identified in the image. The system may determine two-dimensional (2D) coordinates for presentation of an augmented reality object on the display based on the 3D coordinates of the intersection point. The system may superimpose a representation of the augmented reality object on the image based on the 2D coordinates.

A COMPUTER-IMPLEMENTED METHOD FOR MANIPULATING THREE-DIMENSIONAL MODELED OBJECTS OF AN ASSEMBLY IN A THREE-DIMENSIONAL SCENE

A computer-implemented method for manipulating three-dimensional modeled objects of an assembly in a three-dimensional scene, comprising the steps of: - determining at least a first set of at least one object and a second set of at least one object among said three-dimensional modeled objects of the assembly; - grouping the at least one object of the first set in a first three-dimensional bounding box (BB) and the at least one object of the second set in a second three-dimensional bounding box (BB); and relatively reorganizing the bounding boxes (BB).

METHOD FOR CREATING SPATIALLY BALANCED BOUNDING VOLUME HIERARCHIES FOR USE IN A COMPUTER GENERATED DISPLAY OF A COMPLEX STRUCTURE

Disclosed is a method for processing an arbitrary collection of objects, forming a complex structure, into a hierarchy of bounding volumes, from a root volume bounding all objects, to sub-volumes bounding individual objects or assemblies thereof, for use as successive approximations to said objects in a computer generated display. The method includes the first step of creating a bounding volume for each of the objects. Selected bounding volumes are then processed through a combining algorithm determining whether or not, based upon a geometric relationship between the bounding volumes and the higher level, root volume, the selected bounding volumes can be combined. If it is determined that the bounding volumes can be combined, a new bounding volume is created with the combined volumes comprising sub-volumes thereof. This process systematically repeats and attempts to combine all sub-volumes. The combining algorithm preferably allows a combination if the volumes of the combination of the ...

Fast rendering secondary curved surface

PROCESS AND DEVICE OF SIMULATION OF CONTACT BY IMAGES HAVE PLANE DEPTH.

INFORMATION PROCESSING DEVICE

METHOD FOR AUTOMATICALLY SET UP JOINTS TO CREATE FACIAL ANIMATION OF 3D FACE MODEL AND COMPUTER PROGRAM

METHOD FOR CONTROLLING REFRIGERATOR AND REFRIGERATOR CONTROLLER

According to an embodiment, a refrigerator controller comprises: an image acquisition unit receiving a current frame photographing an internal condition of a refrigerator; an area detection unit calculating a pixel value with respect to each pixel of the current frame, and comparing the pixel value with another pixel value with respect to a corresponding pixel of a previous frame to detect a change area; a divided image generating unit generating a divided image changing a pixel value of each pixel of the current frame to a first value or a second value based on the change area; a box extraction unit extracting a bounding box having a pixel set with the first value contained in the divided image; and a determination unit determining a change in a condition of the refrigerator by comparing the bounding box extracted through the box extraction unit with at least one pre-stored bounding box. COPYRIGHT KIPO 2018 (111) Image acquisition unit (112) Buffer (113) Region detection unit (114) Divided ...

SYSTEM AND METHOD FOR IMPROVING IMAGE QUALITY OF OBJECT OF INTEREST

Method and Apparatus for indoor space modeling with function, behavior and structure for intelligent indoor service

ELECTRONIC DEVICE AND OPERATING METHOD THEREOF

According to various embodiments of the present invention, disclosed are a method to select an object by dragging through an electronic device and a device thereof. According to various embodiments of the present invention, the electronic device includes: a display; a memory; and a processor connected functionally with the display and the memory. The processor is able to be configured to display an image through the display, to draw a selecting tool on the image in correspondence with user input, to determine a correction range based on a neighboring object of the selecting tool in the event of the completion of the user input, and to display the selecting tool by correcting the selecting tool as the nearest neighbor to the object based on the correction range. Various embodiments of the present invention are possible. COPYRIGHT KIPO 2018 (410) Exploration area processing module (420) Feature point processing module (430) First feature point processing module (440) Second feature point ...

METHOD FOR GENERATING AND SEARCHING FOR ACCELERATION STRUCTURE

The present invention relates to a method for generating an acceleration structure used for tracking light. The present invention comprises: a step of dividing primitives included in a three-dimensional space into bounding boxes; a step of obtaining information about a location where the bounding boxes overlap with each other; and a step of generating the acceleration structure indicating the inclusion relationship and location information of the bounding boxes. COPYRIGHT KIPO 2017 (610) Node <- Root node (620) Determine an overlap location of child nodes of the node (630) Input overlap location information of the child nodes (640) Node <- Next node (650) Last node? (AA) Start (BB) End ...

METHOD FOR GENERATING AND SEARCHING ACCELERATION STRUCTURE

A method for searching an acceleration structure according to an embodiment of the current invention comprises the steps of: (a) searching a leaf node intersecting with beam data in the acceleration structure; and (b) completing searching when there is no node including overlapped child nodes of parent nodes of the intersected leaf node, and additionally searching overlapped child nodes when there is a node including overlapped child nodes. COPYRIGHT KIPO 2016 (910) Search a leaf node intersecting with beam data in an acceleration structure (920) Determine whether there is a node including overlapped child nodes of parent nodes of the intersected leaf node (930) Additionally search overlapped child nodes (AA) Start (BB) End ...

Multi-space rendering with configurable transformation parameters

Techniques are disclosed relating to rendering graphics objects. In some embodiments, a graphics unit is configured to transform graphics objects from a virtual space into a second space according to different transformation parameters for different portions of the second space. This may result in sampling different portions of the virtual space at different sample rates, which may reduce the number of samples required in various stages of the rendering process. In the disclosed techniques, transformation may occur prior to rasterization and shading, which may further reduce computation and power consumption in a graphics unit, improve image quality as displayed to a user, and/or reduce bandwidth usage or latency of video content on a network. In some embodiments, a transformed image may be viewed through a distortion-compensating lens or resampled prior to display.

GENERAL METHOD FOR DETERMINING A LIST OF ELEMENTS POTENTIALLY VISIBLE BY REGION FOR THREE-DIMENSIONAL LARGE SCENES REPRESENTING VIRTUAL VARIABLE ALTITUDE CITIES

L'invention concerne un procédé de détermination de listes d'objets potentiellement visibles dans une scène 3D comprenant des objets (100) sur un terrain (101) d'altitude variable et dans laquelle un utilisateur est susceptible de se déplacer virtuellement sur une zone de déplacement correspondant à l'ensemble des zones du terrain non recouvertes par les objets (100), caractérisé en ce qu'il comprend les étapes consistant à créer des objets terrain (104) correspondant chacun à une partie du terrain, partitionner la zone de déplacement de l'utilisateur en un réseau de cellules de vue, déterminer, pour chaque cellule de vue, une liste d'objets potentiellement visibles par l'utilisateur depuis la cellule de vue, ladite liste étant déterminée en testant la visibilité des objets (100) de la scène et des objets terrain (104) créés.

Systems, methods, and media for modifying a mesh for a material point method utilized for geotechnical engineering

In an example embodiment, a process may calculate a fill ratio for each element in the background mesh. The process may identify partially filled elements of the background mesh based on the calculated fill ratios. The process may transform each partially filled element to an empty or filled element by identifying a node of the partially filled element and moving the identified node to a different location in the background mesh. The process may generate an updated background mesh, that includes only empty and filled elements, that may be utilized to perform one or more calculations for one or more time steps in a modeling/simulation environment to simulate a behavior of a physical material/object that may exhibit deformations.

Processing systems and methods having a machine learning engine for providing a surface dimension output

Systems and apparatuses for generating surface dimension outputs are provided. The system may collect an image from a mobile device. The system may analyze the image to determine whether they comprise one or more standardized reference objects. Based on analysis of the image and the one or more standardized reference objects, the system may determine a surface dimension output. The system may determine one or more settlement outputs and one or more repair outputs for the driver based on the surface dimension output.

METHODS AND SYSTEMS FOR PERFORMING SLEEPING OBJECT DETECTION IN VIDEO ANALYTICS

Methods, apparatuses, and computer-readable media are provided for maintaining blob trackers for video frames. For example, a blob tracker maintained for a current video frame is identified. The blob tracker is associated with a blob detected in one or more video frames. The blob includes pixels of at least a portion of a foreground object in the one or more video frames. A current bounding region of the blob tracker for the current video frame is compared to a previous bounding region of the blob tracker for a previous video frame that is obtained earlier in time than the current video frame. It can be determined whether the current bounding region has decreased in size as compared to a size of the previous bounding region, and whether a first color characteristic of pixels of the current video frame included in the previous bounding region is within a threshold from a second color characteristic of pixels of the previous video frame included in the previous bounding region. In some examples ...

METHOD FOR DETECTING RAISED PAVEMENT MARKERS, COMPUTER PROGRAM PRODUCT AND CAMERA SYSTEM FOR A VEHICLE

A method is disclosed for detecting raised pavement markers in an environment of a vehicle by a camera system. The method includes capturing at least one first image of at least one first part of the environment by at least one first camera of the camera system and analyzing the at least one first image and determining whether at least one first pavement marker is present in the environment in dependency of a result of the analysis of the at least one first image. The method further includes capturing at least one second image of at least one second part of the environment by at least one second camera of the camera system and analyzing the at least one second image and determining whether the at least one first or at least one second pavement marker is present in the environment in dependency of a result of the analysis.

VISUAL INTENT TRIGGERING FOR VISUAL SEARCH

Representative embodiments disclose mechanisms to perform visual intent classification or visual intent detection or both on an image. Visual intent classification utilizes a trained machine learning model that classifies subjects in the image according to a classification taxonomy. The visual intent classification can be used as a pre-triggering mechanism to initiate further action in order to substantially save processing time. Example further actions include user scenarios, query formulation, user experience enhancement, and so forth. Visual intent detection utilizes a trained machine learning model to identify subjects in an image, place a bounding box around the image, and classify the subject according to the taxonomy. The trained machine learning model utilizes multiple feature detectors, multi-layer predictions, multilabel classifiers, and bounding box regression.

SYSTEMS AND METHODS FOR REDUCING A PLURALITY OF BOUNDING REGIONS

A method performed by an electronic device is described. The method includes generating a plurality of bounding regions based on an image. The method also includes determining a subset of the plurality of bounding regions based on at least one criterion and a selected area in the image. The method further includes processing the image based on the subset of the plurality of bounding regions.

SHADOWS FOR INSERTED CONTENT

Systems and methods for generating shadows for inserted content are provided. The inserted content may include augmented reality content that is inserted into an image of a physical space. An example includes determining a location to insert content within an image. The content may include a polygonal mesh defined in part by a skeleton that has a plurality of joints. Examples may further include selecting a plurality of selected joints form the plurality of joints. Examples may also include generating a shadow polygon based on the content and determining shadow contributions values for the plurality of selected joints for pixels of the shadow polygon. Examples may also include combining the shadow contribution values from the selected joints to generate shadow magnitude values for the pixels, rendering the shadow polygon using the shadow magnitude values, and overlaying the inserted content on the rendered shadow polygon.

Dynamic image capture system

Systems and methods are provided for receiving image data via a camera of a computing device, the image data comprising a plurality of image frames; displaying a 3D reconstruction of the image data on a graphical user interface (GUI) displayed on a computing device as the image data is received and the 3D reconstruction of the image data is generated; detecting at least one object corresponding to one or more of a plurality of predefined object types in the image data; determining dimensions of the at least one object in 3D space based on the 3D reconstruction of the image data; and displaying in the GUI the at least one detected object.

ON-SHELF IMAGE BASED OUT-OF-STOCK DETECTION

An out-of-stock detection system notifies store management that a product is out of stock. The system captures images of a shelf and determines the position product labels thereon. For each product label, a bounding box is generated based on the position of each product label on the shelf. The system then identifies a product for each product label based on information within each product label and, for each product label, stores a product identified for each bounding box. Accordingly, the system performs an out-of-stock detection process that includes capturing additional image data of the shelf periodically that includes each bounding box, providing a portion of the additional image data for each bounding box to a model trained to determine whether the bounding box contains products, sending a notification for a product determined to be out of stock to a store client device based on output from the model.

Apparatus and method of traversing acceleration structure in ray tracing system

A method of traversing an acceleration structure (AS) in a ray tracing system includes obtaining information about child nodes of a target node included in the AS; determining whether each of the child nodes intersects a ray based on the obtained information; determining a next target node among at least one child node that intersects the ray; and performing an operation corresponding to a type of the determined next target node.

Object detection apparatus and method for vehicle

An object detection apparatus includes a first camera unit, a second camera unit, and a control unit. The first camera unit includes one or more cameras, and is configured to capture an image around a vehicle. The second camera unit includes one or more cameras, and is configured to capture an image of an area ahead of the vehicle. The control unit is configured to: determine a displacement of a feature point positioned in a common area from the image acquired via the first camera unit; determine a pixel displacement of the feature point in the image acquired via the second camera unit; and determine distance information to an object recognized in the image captured via the second camera unit based on the displacement of the feature point and the pixel displacement of the feature point.

NEURAL RENDERING USING TRAINED ALBEDO TEXTURES

Methods and systems are disclosed for performing operations for generating a photorealistic rendering of an object. The operations include: accessing a set of albedo textures and a machine learning model associated with a real-world object, the set of albedo textures and a machine learning model having been trained based on a plurality of viewpoints of the real-world object; obtaining a three-dimensional (3D) mesh of the real-world object; receiving input that selects a new viewpoint that differs from the plurality of viewpoints of the real-world object; and generating a photorealistic rendering of the real-world object from the new viewpoint based on the 3D mesh of the real-world object, the set of albedo textures, and the machine learning model associated with the real-world object.

NEURAL NETWORK BASED DETECTION OF IMAGE SPACE SUITABLE FOR OVERLAYING MEDIA CONTENT

Disclosed herein are system, method and computer readable storage medium for detecting space suitable for overlaying media content onto an image. The system receives a candidate image which may be an image or a video frame. The candidate image is then input into a neural network. The neural network may output coordinates and one or more dimensions representing one or more bounding boxes for inserting media content into the candidate image. The one or more bounding boxes may be transmitted with a request for a media content item to be displayed in a bounding box. In response to the request the media content item may be received, and the candidate image and the media content item overlaid on top of the candidate image within the bounding box may be displayed.

Tiling a primitive in a graphics processing system

Tiling a primitive 1102 in a graphics processing system to determine which of a plurality of tiles 1122 of a rendering space the primitive intersects or overlaps with. The tiling involves determining a bounding box for the primitive and, for each boundary 1114 between lines (e.g. rows or columns) of tiles, finding intersection points 1108, 1118 between the boundary and edges of the primitive and using these points to determine which tiles contain at least part of the primitive. Preferably, start 1122s and end 1122e tiles are identified as the respective first or last tiles in a line which contain either vertices or intersection points; and tiles which overlap the primitive in each given line of tiles are determined as all tiles between and including the start and end tiles. Preferably, the lines of tiles are rows if there are fewer rows than columns in the bounding box and the lines are columns of tiles if there are fewer rows than columns. Further independent claims are for tiling primitives ...

Non-parametric method of and system for estimating dimensions of objects of arbitrary shape

A non-parametric method of, and system for, dimensioning an object of arbitrary shape, captures a three-dimensional (3D) point cloud of data points over a field of view containing the object and a base surface on which the object is positioned, detects a base plane indicative of the base surface from the point cloud, extracts the data points of the object from the point cloud, processes the extracted data points of the object to obtain a convex hull, and fits a bounding box of minimum volume to enclose the convex hull. The bounding box has a pair of mutually orthogonal planar faces, and the fitting is performed by orienting one of the faces to be generally perpendicular to the base plane, and by simultaneously orienting the other of the faces to be generally parallel to the base plane.

Clipping of graphics primitives

Techniques for performing clipping of graphics primitives with respect to a clipping boundary are described. The clipping step may be performed separately for each tile of a graphics frame to be rendered, after a primitive list for the tile has been read from a primitive memory. Clipping may be performed only for larger primitives whose size exceeds a given threshold 100. If the primitive size exceeds the threshold a clipping operation is performed, during which new vertex data is generated 106 for the clipped primitive. If the primitive size does not exceed the threshold, the clipping operation is not performed 102. Subsequent processing of the primitives is performed on the outcome of this determination 104,108. Clipping of a primitive to the clipping boundary may be performed inexactly so that only a single clipped primitive is generated which may extend beyond the clipping boundary. A clipped primitive generated by clipping may be used for a depth function calculation of a primitive ...

3-D Graphics rendering with implicit geometry

Systems and methods for a 3-D scene acceleration structure creation and updating

Improvements to a tiling system for 3D rendered graphics

Techniques for interior coordinates

A method for determining interior coordinates is disclosed. The method includes receiving information specifying an object having a plurality of sites and a boundary. Interior coordinates associated with each of the plurality of sites are determined based on the boundary. The interior coordinate associated with each of the plurality of sites represent a system of coordinates that satisfy several properties, including non-negativity and interior locality. At least one value associated with the plurality of sites is then interpolated using the interior coordinates.

Reorientation of cardiac PET images

Generation and management of winding numbers in graphic processing

Non-parametric method of and system for estimating dimensions of objects of arbitrary shape

System and methods for efficient vehicle extent estimation

A device configured to: generate a bounding box enclosing a cluster of points corresponding to a vehicle, sort the points into a first set of points located within a distance from a first side of the box and a second set of points located within the distance from the second side; generate convex hulls based on the sets of points, determine intervals of orientations associated with the bounding box within which the sets of points remain constant (i.e. remain within the same convex hull); determine an orientation of the bounding box based on the intervals of orientations that minimizes distances from the first and second set of points to the first and second sides respectively; and control operation of an autonomous vehicle based on the determined orientation. The orientations may be determined through rotating the bounding box around the clusters. The method reduces the calculations required to determine the best fit for a bounding box around a cluster of points.

Artificial image generation for training an object detection system

A device is disclosed. The device may obtain a machine model of a machine and associated motion data for the machine model of the machine. The device may render the machine model into a rendered environment associated with a set of characteristics. The device may capture a set of images of the machine model and the rendered environment based on rendering the machine model into the rendered environment. The device may determine bounding box data for the set of images of the machine model and the rendered environment based on the position of the machine model within the rendered environment relative to an image capture orientation within the rendered environment. The device may provide the set of images of the machine model and the rendered environment and the bounding box data as a data set for object detection.

Apparatus for multi-angle screen coverage analysis

Embodiments provide multi-angle screen coverage analysis. In some embodiments, a system obtains a computer graphics generated image having at least one target object for analysis. The system determines screen coverage information and depth information for the at least one target object. The system then determines an asset detail level for the at least one target object based on the screen coverage information and the depth information. The system then stores the asset detail level in a database, and makes the asset detail level available to users.

TRIANGLE RENDERING USING DIRECT EVALUATION

This disclosure describes an apparatus, such as a wireless communication device, that applies a direct evaluation technique to render triangles for the 3D graphical environment. The apparatus includes a rendering engine that defines a rectangular area of pixels, referred to as a bounding box, that bounds the area to be rendered. The rendering engine evaluates coordinates associated with the pixels of the rectangular area to selectively render those pixels that fall within the triangular area. The direct evaluation triangle rendering algorithm may require fewer complex operations than the more computationally intensive interpolation process employed by other systems. As a result, the apparatus may present a 3D graphical environment while preserving as much as possible the available power.

FACE DETECTION USING SMALL-SCALE CONVOLUTIONAL NEURAL NETWORK (CNN) MODULES FOR EMBEDDED SYSTEMS

Embodiments described herein provide various examples of a face detection system, based on using a small-scale hardware convolutional neutral network (CNN) module configured into a multi-task cascaded CNN. In some embodiments, a subimage-based CNN system can be configured to be equivalent to a large-scale CNN that processes the entire input image without partitioning such that the output of the subimage-based CNN system can be exactly identical to the output of the large-scale CNN. Based on this observation, some embodiments of this patent disclosure make use of the subimage-based CNN system and technique on one or more stages of a cascaded CNN or a multitask cascaded CNN (MTCNN) so that a larger input image to a given stage of the cascaded CNN or the MTCNN can be partitioned into a set of subimages of a smaller size. As a result, each stage of the cascaded CNN or the MTCNN can use the same small-scale hardware CNN module that is associated with a maximum input image size constraint.

THREE-DIMENSIONAL ANNOTATIONS FOR STREET VIEW DATA

The present invention relates to annotating images. In an embodiment, the present invention enables users to create annotations corresponding to three-dimensional objects while viewing two-dimensional images. In one embodiment, this is achieved by projecting a selecting object onto a three-dimensional model created from a plurality of two-dimensional images. The selecting object is input by a user while viewing a first image corresponding to a portion of the three-dimensional model. A location corre-sponding to the projection on the three-dimensional model is determined, and content entered by the user while viewing the first image is associated with the location. The content is stored together with the location information to form an annotation. The anno-tation can be retrieved and displayed together with other images corresponding to the location.

SELECTION OF A VIEWPOINT OF A SET OF OBJECTS

It is proposed a computer-implemented method for selecting a viewpoint of a set of objects in a bounding box. The bounding box comprising at least one volume wherein the set of objects is enclosed. The method comprises identifying faces of the at least one volume; associating with each identified face a viewpoint on the set of object; selecting, upon user action, one of the faces associated with one viewpoint; and displaying a view of the set of objects according to the viewpoint associated with the selected face.

METHOD FOR REPRESENTING A PARTICIPATING MEDIA IN A SCENE AND CORRESPONDING DEVICE

The invention relates to a method for rendering a participating media (10) delimited by a bounding box (30) and rendered from a viewpoint (14), the media (10) being at a determined distance from the viewpoint according to a viewing direction. In order to represent the limits of the participating media, the method comprises steps of: - for at least one point (300) of the volume formed by said bounding box (30), estimation of a set of directions (31 to 3i) having for origin said at least one point (300): - for each direction (31 to 3i), estimation of a first intersection point (31m), corresponding to the intersection between said direction and said participating media (10), for which the associated density value is greater than a first threshold value, and - estimation of third intersection points, corresponding to the intersections between said viewing direction and said participating media, from an item of information representative of distances separating the first intersection point ( ...

SYSTEMS AND METHODS FOR CREATING A THREE-DIMENSIONAL TEXTURE ATLAS

Systems and methods for reducing the amount of texture cache memory needed to store a texture atlas by using uniquely grouped refined triangles to create each texture atlas.

PROCESS AND SYSTEM OF RETURNED Of a VIRTUAL SCENE IN THREE DIMENSIONS

La présente invention concerne un procédé de rendu d'une scène virtuelle en trois dimensions comprenant: une première phase de préparation de données comprenant au moins une étape de regroupement de données de la scène dans une structure de données (OD) correspondant à une division spatiale de la scène en une grille régulière de façon à former des métavoxels, les données numériques étant stockées dans la structure de donnée optimisée (OD) sous différents niveaux de détail par métavoxel, les différents niveaux correspondant à des simplifications géométriques, et une seconde phase de rendu temps réel utilisant la structure de données géométriques (OD) pour le rendu d'au moins une image comprenant une étape de détermination des niveaux de détails à utiliser pour chaque métavoxel prenant en compte une limite de budget en mémoire d'un ordinateur sur lequel est exécuté au moins la phase de rendu et effectuant une simulation de l'occupation en mémoire des métavoxels à charger suivant leur niveau ...

SYSTEMS AND METHODS FOR DETECTING OBJECTS

APPARATUS AND METHOD FOR RECOGNIZING CHARACTER

PLENOPTIC IMAGE PROCESSING APPARATUS, SYSTEM HAVING THE SAME, AND OBJECT SEGMENTATION METHOD THEREOF

A METHOD FOR PIG SEGMENTATION USING CONNECTED COMPONENT ANALYSIS AND YOLO ALGORITHM

APPARATUS AND METHOD FOR SEARCHING ACCELERATION STRUCTURE IN RAY TRACING SYSTEM

The present invention relates to an apparatus and a method for searching an acceleration structure in a ray tracing system. The method for searching an acceleration structure in a ray tracing system comprises: acquiring information of a plurality of child nodes on a target node included in an acceleration structure; determining whether each of the child nodes intersects with a ray, based on the acquired information; determining a next target node from at least one child node intersecting with the ray; and performing an operation corresponding to a type of the determined next target node, thereby reducing the time spent on searching the acceleration structure. COPYRIGHT KIPO 2016 (300) Apparatus for searching an acceleration structure (310) Information acquisition unit (320) Cross examination unit (330) Control unit ...

영상 데이터에 포함된 얼굴 및 신체 비식별화 방법 및 이를 수행하기 위한 장치

... 영상 데이터에 포함된 얼굴 및 신체 비식별화 방법 및 이를 수행하기 위한 장치가 개시된다. 다양한 실시예에 따른 얼굴 및 신체 정보를 비식별화하는 비식별화 방법은 입력 이미지에 포함된 얼굴 및 신체 중 적어도 어느 하나의 경계부 박스(bounding box)를 검출하는 동작, 상기 입력 이미지에서 상기 경계 박스에 포함된 상기 적어도 어느 하나의 경계선을 따라 추출(예: 세그멘테이션(segmentation))하여 제1 입력 데이터를 획득하는 동작, 상기 제1 입력 데이터에서 상기 적어도 어느 하나를 구성하는 요소의 특징 및 각도를 알 수 있는 랜드마크를 추출하여 제2 입력 데이터를 획득하는 동작과 상기 제1 입력 데이터 및 상기 제2 입력 데이터를 비식별화 모델에 입력시켜 상기 입력 이미지에서 상기 적어도 어느 하나가 비식별화된 이미지를 생성하는 동작을 포함할 수 있다.

SYSTEM AND METHOD FOR COMPONENT SCATTERING FROM A LIST

A system, method, and computer program for component scattering from a bill of material, comprising selecting a plurality of components from a list; calculating a bounding box for each of said plurality of components; calculating a centroid corresponding to each of said bounding boxes; calculating a scatter circle for placement of said plurality of components; and calculating a number of locations to place said plurality of components on said scatter circle; whereby said plurality of components are added to an assembly view in a single operation such that a largest of said plurality of components is at a start point and appropriate means and computer-readable instructions.

SYSTEM AND METHOD FOR JOINT IMAGE AND LIDAR ANNOTATION AND CALIBRATION

A calibration system includes a processing circuit configured to receive data corresponding to a first image where the data includes information corresponding to a 2D BEV of a scene, and to receive data corresponding to a second image where the data includes information corresponding to a 2D image of the scene. The processing circuit may determine a salient feature in the first image or the second image, and project, based on a first calibration setting, the salient feature from the first image to the second image or vice versa. The processing circuit may select a region corresponding to the projected salient feature in the first image or the second image, determine an identity of the salient feature, and determine a second calibration setting based on the identity of the salient feature and the selected region corresponding to the salient feature in the 2D BEV and the 2D image.

Isolation of aneurysm and parent vessel in volumetric image data

A framework for isolating an aneurysm and parent vessel in volumetric image data is provided herein. In accordance with one aspect, the framework generates a refined mask by performing region growing starting at an aneurysm dome point to eliminate vessels that are indirectly connected to an aneurysm or parent vessel in volumetric image data. A final mask may be generated based at least in part on the refined mask by eliminating any kissing vessel connected with the aneurysm from the refined mask. The final mask may then be used for segmentation of the aneurysm and the parent vessel in the volumetric image data.

METHOD AND APPARATUS FOR GENERATING LEARNING DATA REQUIRED TO LEARN ANIMATION CHARACTERS BASED ON DEEP LEARNING

Disclosed are a learning data generation method and apparatus needed to learn animation characters on the basis of deep learning. The learning data generation method needed to learn animation characters on the basis of deep learning may include collecting various images from an external source using wired/wireless communication, acquiring character images from the collected images using a character detection module, clustering the acquired character images, selecting learning data from among the clustered images, and inputting the selected learning data to an artificial neural network for character recognition.

UTILIZING OBJECT ATTRIBUTE DETECTION MODELS TO AUTOMATICALLY SELECT INSTANCES OF DETECTED OBJECTS IN IMAGES

The present disclosure relates to an object selection system that accurately detects and automatically selects target instances of user-requested objects (e.g., a query object instance) in a digital image. In one or more embodiments, the object selection system can analyze one or more user inputs to determine an optimal object attribute detection model from multiple specialized and generalized object attribute models. Additionally, the object selection system can utilize the selected object attribute model to detect and select one or more target instances of a query object in an image, where the image includes multiple instances of the query object. 1. A non-transitory computer-readable medium storing instructions that , when executed by at least one processor , cause a computing device to:identify, from a selection query corresponding to a digital image, a query object and a corresponding object attribute;detect multiple instances of the query object in the digital image;determine an object attribute detection model from a plurality of object attribute detection models based on analyzing the object attribute;detect that a first query object instance from the multiple query object instances is a target query object instance based on the object attribute and the determined object attribute detection model; andprovide the digital image with the target query object instance selected in response to the selection query.2. The non-transitory computer-readable medium of claim 1 , further comprising instructions that claim 1 , when executed by the at least one processor claim 1 , cause the computing device to parse the selection query to determine:a noun that identifies the query object; andan adjective that identifies object attribute.3. The non-transitory computer-readable medium of claim 1 , further comprising instructions that claim 1 , when executed by the at least one processor claim 1 , cause the computing device to:analyze the object attribute to identify an ...

Generation of a Control Stream for a Tile

A method of processing primitives within a tiling unit of a graphics processing system is described. The method comprises determining whether a primitive falls within a tile based on positions of samples within each pixel. If it is determined that the primitive does fall within a tile based on the positions of samples within pixels in a tile, an association between the tile and the primitive is stored to indicate that the primitive is present in the tile. For example, an identifier for the primitive may be added to a control stream for the tile to indicate that the primitive is present in the tile. Various different methods are described to make the determination and these may be used separately or in any combination.

Techniques for representing 3D scenes using fixed point data

A quantization transform, corresponding to a geometric object, is determined. The geometric object represents at least a portion of an object in a three-dimensional scene. The quantization transform is suitable for converting a floating point space to a fixed point space, where the floating point space contains floating point data corresponding to the geometric object. The quantization transform is used to convert floating point data to fixed point data. The floating point data is typically vertices corresponding to the geometric object. Additionally, a quantization transform corresponding to a geometric object is determined. The geometric object represents at least a portion of the three-dimensional scene. The quantization transform is suitable for converting a floating point space to a fixed point space, and the fixed point space contains one or more fixed point data corresponding to the geometric object. The floating point space defines at least the portion of the three-dimensional scene ...

AUTOMATED TRAINING DATA COLLECTION FOR OBJECT DETECTION

A method, system, and computer program product for automated collection of training data and training object detection models is provided. The method generates a set of reference images for a first set of products. Based on the set of reference images, the method identifies a subset of products within an image stream. Based on the subset of products, a second set of products is determined within the image stream. The method identifies a set of product gaps based on the subset of products and the second set of products. The method generates a product detection model based on the set of reference images, the subset of products, the second set of products, and the product gaps.

METHOD AND DEVICE FOR DETERMINING PLURALITY OF LAYERS OF BOUNDING BOXES, COLLISION DETECTION METHOD AND DEVICE, AND MOTION CONTROL METHOD AND DEVICE

A method for determining a plurality of layers of bounding boxes of an object includes determining a polyhedron capable of accommodating the object therein as a first-layer bounding box. The method also includes selecting one vertex from a plurality of vertices of the first-layer bounding box as a target vertex, and determining, by processing circuitry of a computing device, a support plane of the object. The support plane has a normal vector that has a specific direction corresponding to the target vertex and that is closest to the target vertex, the support plane being a plane passing through a point on a surface of the object such that the object is completely located on one side of the support plane. The method further includes cutting the first-layer bounding box based on at least the support plane to form a smaller bounding box, as a second-layer bounding box.

METHODS AND SYSTEMS FOR PROCESSING IMAGE DATA

A computer-implemented system for processing image data. The system comprises: a sensor operable to capture image data comprising an image of an environment of the sensor; and a processing circuit comprising: a processor; and a computer-readable storage medium comprising computer-readable instructions which, when executed, cause the processor to: receive the image data from the sensor; process the image data using a first path through a neural network to obtain first data, the first path configured to indicate a presence in the environment of one or more objects of a predetermined object type; process the image data using a second path through a neural network to obtain second data, the second path configured to indicate a presence in the environment of one or more object characteristics corresponding to the predetermined object type; and generate output data using the first and second data, wherein the first and second paths are arranged to be enable the first and second data to be obtained ...

MEDICAL IMAGE DATA

Medical image data is received at a data processing system, which is an artificial intelligence-based system. An identification process is performed at the data processing system to identify a subset of the medical image data representing a region of interest including one or more target tendons. A determination process is performed at the data processing system to determine one or more characteristics relating to one or more abnormalities of the one or more target tendons. Abnormality data is output, the abnormality data relating to the one or more abnormalities and being based on the one or more characteristics.

B-rep matching for maintaining associativity across CAD interoperation

A method allows transferring references from vertices, edges, and faces of a solid or surface model to an edited model, by finding correspondences between vertices, edges, faces, and bodies of the models. The method begins by quickly finding matching bodies and aligning them, considering factors such as size, complexity, and the location of its vertices, edges, and faces. For every pair of matched bodies, it then matches exactly coincident vertices, edges, and faces. Of the remaining unmatched topology, it matches edges and faces in the new model that overlap those of the old model considerably. It uses near-neighbor structures to speed up the search for correspondences by eliminating old-new pairs of vertices, edges, and faces that do not overlap. Finally, it iteratively matches unmatched vertices, edges, and faces whose adjacent vertices, edges, and faces are in exact correspondence between the old and new models.

DEEP LEARNING SYSTEM FOR CUBOID DETECTION

... 32259646 ...

Method and system for evaluating an object detection model

Disclosed herein are computer-implemented methods, computer-implemented systems, and non-transitory, computer-readable media for evaluating performance of an object detection model. One computer-implemented method includes generating a predicted bounding box representing an object based on the object detection model, where the object is positioned proximate to one or more adjacent objects. An area of intersection is determined between the predicted bounding box and a groundtruth bounding box of the object. A modified area of union is determined between the predicted bounding box and the groundtruth bounding box of the object. A score equal to the area of intersection between the predicted bounding box and the groundtruth bounding box of the object divided by the modified area of union is determined, where the score represents the performance of the object detection model.

FINGERPRINT PREVIEW QUALITY AND SEGMENTATION

A ridge flow based fingerprint image quality determination can be achieved independent of image resolution, can be processed in real-time and includes segmentation, such as fingertip segmentation, therefore providing image quality assessment for individual fingertips within a four finger flat, dual thumb, or whole hand image. A fingerprint quality module receives from one or more scan devices ridge-flow-containing imagery which can then be assessed for one or more of quality, handedness, historical information analysis and the assignment of bounding boxes.

Distance field coupled fitted deformation lattices for shape modification

A method and system provide the ability to modify a three-dimensional (3D) model in a shape editing system. The 3D model is obtained and faces of the model are selected as features (S). A subset (S′) of the model that are fixed are selected. Shape modification operations to be performed are prescribed. A deformation lattice is constructed by setting up a lattice structure with control points. Parametric space (u,v,w) is defined in terms of vertices of the lattice structure. Euclidean space (x,y,z) of the 3D model is mapped to the parametric space (u,v,w). The deformation lattice is evaluated by selecting control points, and either affine transformations are applied directly to the selected control points, or the deformation lattice is deformed based on a discrete fitting problem. The evaluated deformed model is then output.

Ray tracing hierarchy

A hierarchy of bounding volumes for objects is generated for use during ray tracing. The hierarchy of bounding volumes improves image quality and well as processing time during rendering. A memory cost factor is identified that relates to the increase in the amount of memory for a subdivided volume compared to its undivided form. An estimate is made of the average number of surface intersections made with a ray passing through the undivided volume and the divided volume, and a factor is evaluated that measures the reduction in such intersections resulting from the subdivisions. A comparison between the memory cost factor and the intersection reduction factor determines whether or not a bounding volume is further subdivided. These tests are then applied recursively to the newly created bounding volumes.

Full Body Virtual Reality Utilizing Computer Vision From a Single Camera and Associated Systems and Methods

Methods and systems for constructing a three-dimensional (3D) model of a user in a virtual environment for full body virtual reality (VR) applications are described. The method includes receiving an image of the user captured using an RGB camera; detecting a body bounding box associated with the user using a first trained neural network; determining a segmentation map of the user, based on the body bounding box; determining a two-dimensional (2D) contour of the user from the segmentation map; forming a 3D extrusion model by extruding the 2D contour; and constructing the 3D model of the user in the virtual environment by applying a geometric transformation to the 3D extrusion model. Applications of full body VR include physical training and fitness sessions, games, control of computing devices, manipulation and display of data, interactive social media with VR, and the like.

Data integration from multiple sensors

Disclosed are methods and devices related to autonomous driving. In one aspect, a method is disclosed. The method includes determining three-dimensional bounding indicators for one or more first objects in road target information captured by a light detection and ranging (LIDAR) sensor; determining camera bounding indicators for one or more second objects in road image information captured by a camera sensor; processing the road image information to generate a camera matrix; determining projected bounding indicators from the camera matrix and the three-dimensional bounding indicators; determining, from the projected bounding indicators and the camera bounding indicators, associations between the one or more first objects and the one or more second objects to generate combined target information; and applying, by the autonomous driving system, the combined target information to produce a vehicle control signal.

PROCESSING SYSTEM HAVING A MACHINE LEARNING ENGINE FOR PROVIDING A SURFACE DIMENSION OUTPUT

Systems and apparatuses for generating surface dimension outputs are provided. The system may collect an image from a mobile device. The system may analyze the image to determine whether they comprise one or more standardized reference objects. Based on analysis of the image and the one or more standardized reference objects, the system may determine a surface dimension output. The system may determine one or more settlement outputs and one or more repair outputs for the driver based on the surface dimension output.

Method and system for detecting and analyzing objects

A method for detecting objects and labeling the objects with distances in an image includes steps of: obtaining a thermal image from a thermal camera, an RGB image from an RGB camera, and radar information from an mmWave radar; adjusting the thermal image based on the RGB image to generate an adjusted thermal image, and generating a fused image based on the RGB image and the adjusted thermal image; generating a second fused image based on the fused image and the radar information; detecting objects in the images, and generating, based on the fused image, another fused image including bounding boxes marking the objects; and determining motion parameters of the objects.

TRANSFERRING DATA FROM AUTONOMOUS VEHICLES

A system includes at least one imaging sensor and a processor. The processor is configured to acquire detected data describing an environment of an autonomous vehicle using the imaging sensor; derive reference data which describes the environment from a predefined map; compute difference data representing a difference between the detected data and the reference data; and transfer the difference data, wherein an image computed based on the difference data and the reference data represents the detected data. Other embodiments are also described.

Prediction on top-down scenes based on action data

Techniques for determining predictions on a top-down representation of an environment based on vehicle action(s) are discussed herein. Sensors of a first vehicle (such as an autonomous vehicle) can capture sensor data of an environment, which may include object(s) separate from the first vehicle (e.g., a vehicle or a pedestrian). A multi-channel image representing a top-down view of the object(s) and the environment can be generated based on the sensor data, map data, and/or action data. Environmental data (object extents, velocities, lane positions, crosswalks, etc.) can be encoded in the image. Action data can represent a target lane, trajectory, etc. of the first vehicle. Multiple images can be generated representing the environment over time and input into a prediction system configured to output prediction probabilities associated with possible locations of the object(s) in the future, which may be based on the actions of the autonomous vehicle.

METHOD OF GROUPING USER INTERFACES IN AN ENVIRONMENT

Methods for displaying and organizing user interfaces in a computer-generated environment provide for an efficient and intuitive user experience. In some embodiments, user interfaces can be grouped together into a container. In some embodiments, user interfaces can be added to a container, removed from a container, or moved from one location in the container to another location. In some embodiments, a visual indication is displayed before a user interface is added to a container. In some embodiments, a user interface can replace an existing user interface in a container. In some embodiments, when moving a user interface in a computer-generated environment, the transparency of a user interface that is obscured can be modified.

THREE-DIMENSIONAL PERSPECTIVE CORRECTION

Embodiments of the present disclosure set forth techniques for three-dimensional perspective correction, where the techniques include receiving a first image that depicts an object from a first perspective along a first axis; determining a location of the object along the first axis; determining, based on the first image and the location of the object along the first axis, at least one dimension of the object; and adjusting, based on the at least one dimension of the object, a second image that depicts the object from a second perspective along a second axis. In some embodiments, adjusting the second image comprises skewing at least a portion of the second image to depict the object from a third perspective along a third axis.

Three-Dimensional Annotations For Street View Data

The present invention relates to annotating images. In an embodiment, the present invention enables users to create annotations corresponding to three-dimensional objects while viewing two-dimensional images. In one embodiment, this is achieved by projecting a selecting object onto a three-dimensional model created from a plurality of two-dimensional images. The selecting object is input by a user while viewing a first image corresponding to a portion of the three-dimensional model. A location corresponding to the projection on the three-dimensional model is determined, and content entered by the user while viewing the first image is associated with the location. The content is stored together with the location information to form an annotation. The annotation can be retrieved and displayed together with other images corresponding to the location.

Method, Apparatus and System for a Graphical User Interface

A computer-implemented method of positioning a set of objects for display on a user interface, the user interface being generated on a surface of a three-dimensional structure, each member of the set of the objects having a size in at least two dimensions, at least some of which dimensions vary between the objects, is provided. The method includes assigning the set of objects to a two-dimensional portion having a predetermined size, for each as yet unpositioned object in the set of objects, performing a positioning process, the positioning process comprising identifying object occupancy in the portion and allocating the object to a position within the portion on the basis of a largest dimension of the object and unoccupied space in the portion, and mapping the allocated positions to a corresponding coordinate on the surface of the three-dimensional structure.

System and method for constructing a bounding volume hierarchical structure

A system and method for constructing a bounding volume hierarchical structure are disclosed. The method includes defining a parent node for the bounding volume hierarchical structure, the parent node including a parent node bounding volume enclosing a plurality of objects. A first cost is computed for performing an object partition of the parent node bounding volume to produce a first plurality of child node bounding volumes, and a second cost is also computed for performing a spatial partitioning of the parent node bounding volume to produce a second plurality of child node bounding volumes. The bounding volume hierarchical structure is constructed employing the second plurality of child node bounding volumes produced from the spatial partitioning of the parent node bounding volume if the second cost is lower than the first cost.

3d body modeling from one or more depth cameras in the presence of articulated motion

The present disclosure describes systems and techniques relating to generating three dimensional (3D) models from range sensor data. According to an aspect, multiple 3D point clouds, which are captured using one or more 3D cameras, are obtained. At least two of the 3D point clouds correspond to different positions of a body relative to at least a single one of the one or more 3D cameras. Two or more of the 3D point clouds are identified as corresponding to two or more predefined poses, and a segmented representation of the body is generated, in accordance with a 3D part-based volumetric model including cylindrical representations, based on the two 3D point clouds identified as corresponding to the two predefined pose.

Creation and manipulation of hand drawn objects with automatic grouping

Providing a graphical drawing includes maintaining recently drawn strokes in a wet state after the strokes are drawn, causing strokes in a wet state to transition to a dry state based on passage of time, proximity of other recently drawn strokes, and/or explicit user input directing that at least some wet strokes be transitioned to the dry state, and grouping strokes in the dry state for future user manipulation based at least in part on how the strokes transitioned from the wet state to the dry state. Wet ink strokes may transition from the wet state to the dry state in response to not adding any new wet ink strokes for a predetermined amount of time. The predetermined amount of time may be twelve hundred milliseconds.

METHODS AND APPARATUS TO GENERATE SPATIAL AUDIO BASED ON COMPUTER VISION

Methods, apparatus, systems, and articles of manufacture are disclosed to generate spatial audio based on computer vision. An example apparatus includes at least one memory, instructions in the apparatus, and processor circuitry to execute the instructions to determine a position of an audio source based on an image generated via a camera, and apply an audio spatialization filter to an audio signal generated by a microphone based on the position of the audio source. 1. An apparatus comprising:at least one memory;instructions in the apparatus; and determine a position of an audio source based on an image generated via a camera; and', 'apply an audio spatialization filter to an audio signal generated by a microphone based on the position of the audio source., 'processor circuitry to execute the instructions to2. The apparatus of claim 1 , wherein the processor circuitry is to determine the position of the audio source as an azimuth angle.3. The apparatus of claim 1 , wherein the audio source is a person speaking.4. The apparatus of claim 3 , wherein the processor circuitry is to:detect a face of the person speaking in the image; andgenerate a bounding box of the face of the person speaking in the image.5. The apparatus of claim 4 , wherein the processor circuitry is to:determine a lateral location within the image of the person speaking based on a location of the bounding box; anddetermine a proximity of the person speaking with respect to the camera based on a size of the bounding box.6. The apparatus of claim 5 , wherein the processor circuitry is to determine the position of the person speaking based on the lateral location and the proximity.7. The apparatus of claim 5 , wherein the processor circuitry is to determine the proximity of the person speaking to be on a near plane claim 5 , a medium plane claim 5 , or a far plane.8. The apparatus of claim 1 , wherein the microphone is located at the audio source.9. The apparatus of claim 1 , wherein the audio ...

System and method of three-dimensional model generation

A method includes generating, at an electronic device, a three-dimensional model of an object based on a sequence of images captured by an image capture device associated with the electronic device. The method further includes displaying the three-dimensional model via a display device associated with the electronic device. The method also includes, based on detecting that the three-dimensional model includes an anomaly, presenting, via the display device, one or more selectable options to enable correction of the anomaly.

METHOD AND APPARATUS FOR REPRESENTING CORDINATE VALUES OF BOUNDING BOX OF OBJECT

Provided are a method and apparatus for representing coordinate values of a bounding box of an object. The apparatus and corresponding method receive a minimum value and a maximum value of coordinate values of a bounding box as floating point values. A difference value is determined between the minimum and maximum values, and the difference value is converted to an exponent and a mantissa. 1. A method , comprising:receiving a minimum value and a maximum value of coordinate values of a bounding box as floating point values;determining a difference value between the minimum and maximum values; andconverting the difference value to an exponent and a mantissa.2. The method of claim 1 , further comprising:rounding down the minimum value; androunding up the difference value,wherein the determining of the difference value comprises obtaining a difference value between the rounded down minimum value and the maximum value.3. The method of claim 2 , further comprising:pre-scaling the minimum and maximum values to a pre-set first decimal place; andconverting the rounded down minimum value to an integer,wherein the rounding down comprises rounding down the pre-scaled minimum value, andthe determining of the difference value comprises obtaining a difference value between the pre-scaled maximum value and the rounded down minimum value.4. The method of claim 1 , further comprising:rounding up the received maximum value; androunding up the difference value,wherein the determining of the difference value comprises obtaining a difference value between the rounded up maximum value and the received maximum value.5. The method of claim 4 , further comprising:pre-scaling the received minimum and maximum values to a pre-set first decimal place; andconverting the rounded up maximum value to an integer,wherein the rounding up of the received maximum value comprises rounding up the pre-scaled maximum value, and the determining of the difference value comprises obtaining a difference value ...

Pattern Matching Using Edge-Driven Dissected Rectangles

Aspects of the disclosed technology relate to techniques of pattern matching. Matching rectangles in a layout design that match rectangle members of a search pattern are identified based on edge operations. The rectangle members comprise an origin rectangle member and one or more reference rectangle members. Grid element identification values are attached to the matching rectangles. The matching rectangles that match the one or more reference rectangle members in neighborhoods of the matching rectangles that match the origin rectangle member are then analyzed. The neighborhoods are determined based on the grid element identification values. Based on the analysis, matching patterns in the layout design that match the search pattern are determined. 1. One or more computer-readable media storing computer-executable instructions for causing one or more processors to perform a method , the method comprising:identifying matching rectangles in a layout design that match rectangle members of a search pattern based on edge operations and a first set of criteria, the rectangle members comprising an origin rectangle member and one or more reference rectangle members;attaching grid element identification values to the matching rectangles, the grid element identification values being associated with a regular grid that divides the layout design into a number of equal rectangle regions (grid elements);determining matching patterns in the layout design that match the search pattern based on a second set of criteria, the determining comprising analyzing the matching rectangles that match the one or more reference rectangle members in neighborhoods of the matching rectangles that match the origin rectangle member, wherein the neighborhoods are determined based on the grid element identification values; andstoring the matching patterns.2. The one or more computer-readable media recited in claim 1 , wherein the edge operations are edge-based operations of a design rule checking (DRC) ...

PART FEATURING A BUMP

A computer-implemented method for designing, with a CAD system, a 3D modeled object representing a mechanical part. The method includes displaying a B-rep having faces and edges and representing at least a part of the mechanical part, selecting, by graphical user interaction, at least one face of the B-rep, and, automatically by the CAD system, recognizing a set of faces including the at least one face and representing a bump. The recognizing includes determining a first set of faces representing a depression and a second set of faces representing a protrusion, and selecting, among the first set of faces and the second set of faces, a set of faces having a shortest boundary curve perimeter and being different from both the selected at least one face and the B-rep. The first set of faces and the second set of faces both comprise the at least one face. 1. A computer-implemented method for designing , with a CAD system , a 3D modeled object representing a mechanical part , the method comprising:displaying a B-rep representing at least a part of the mechanical part, the B-rep having faces and edges;selecting, by graphical user interaction, at least one face of the B-rep; and determining a first set of faces and a second set of faces, the first set of faces and the second set of faces both including the at least one face, the first set of faces representing a depression and the second set of faces representing a protrusion, and', 'selecting, among the first set of faces and the second set of faces, a set of faces having a shortest boundary curve perimeter and being different from both the selected at least one face and the B-rep., 'automatically by the CAD system, recognizing a set of faces including the selected at least one face, the set of faces representing a bump, the recognizing of the set of faces including2. The method of claim 1 , wherein:the first set of faces includes a first connected set of one or more faces containing no convex internal edge and no round ...

PREDICTION ON TOP-DOWN SCENES BASED ON ACTION DATA

Techniques for determining predictions on a top-down representation of an environment based on vehicle action(s) are discussed herein. Sensors of a first vehicle (such as an autonomous vehicle) can capture sensor data of an environment, which may include object(s) separate from the first vehicle (e.g., a vehicle or a pedestrian). A multi-channel image representing a top-down view of the object(s) and the environment can be generated based on the sensor data, map data, and/or action data. Environmental data (object extents, velocities, lane positions, crosswalks, etc.) can be encoded in the image. Action data can represent a target lane, trajectory, etc. of the first vehicle. Multiple images can be generated representing the environment over time and input into a prediction system configured to output prediction probabilities associated with possible locations of the object(s) in the future, which may be based on the actions of the autonomous vehicle. 1. A system comprising:one or more processors; and receiving sensor data of an environment captured by a sensor of an autonomous vehicle;', 'receiving action data associated with a candidate action for the autonomous vehicle to perform in the environment, wherein the candidate action comprises one or more of a stay-in-lane action, a turn action, or a lane change action, and wherein the action data is indicative of the candidate action;', 'generating, based at least in part on the sensor data and the action data, a multi-channel image representing a top-down view of the environment, the multi-channel image representing a bounding box associated with a vehicle in the environment proximate the autonomous vehicle, one or more of kinematic information or semantic information associated with the vehicle, and the candidate action associated with the autonomous vehicle;', 'inputting the multi-channel image into a machine learned model trained to generate a heat map comprising a prediction probability of a possible location ...

DRIVING STATE DETECTION METHOD AND APPARATUS, DRIVER MONITORING SYSTEM AND VEHICLE

Embodiments of the present disclosure provide a driving state detection method and apparatus, a driver monitoring system, and a vehicle. The driving state detection method includes: performing head pose detection and eye state detection on a driver image to obtain head pose information and eye state information; and determining detection results of fatigue state and distraction state of the driver according to the head pose information and the eye state information. 1. A driving state detection method , comprising:performing head pose detection and eye state detection on a driver image to obtain head pose information and eye state information;determining whether a deviation angle of a head location of a driver exceeds a preset range according to head location information of the driver in the driver image determined based on the head pose information;in response to determining that the deviation angle of the head location of the driver exceeds the preset range, determining a detection result of distraction state of the driver according to the head pose information and the eye state information; andin response to determining that the deviation angle of the head location of the driver does not exceed the preset range, determining detection results of fatigue state and distraction state of the driver according to the head pose information and the eye state information.2. The method according to claim 1 , wherein determining the detection results of fatigue state and distraction state of the driver according to the head pose information and the eye state information comprises:determining a parameter value of an index for representing the fatigue state of the driver and a parameter value of an index for representing the distraction state of the driver according to the head pose information and the eye state information;determining the detection result of fatigue state of the driver according to the parameter value of the index for representing the fatigue state of the ...

METHOD FOR RECOGNIZING TABLE AND FLOWCHART IN DOCUMENT IMAGES

A table and flowchart detection method is disclosed. First, based on connected component analysis and the sizes of the connected components, a target connected component that corresponds to possible elements of table or flowchart is detected in the input image. The target connected component is broken into corners and edges that connect the corners. Based on the relationship between the corners and edges, it is determined whether the target connected component is a table or a flowchart. For table detection, the edges and corners are linked into horizontal sets and vertical sets, and based on corner counts in the horizontal sets and vertical sets, it is determined whether the target connected component is a table. For flowchart detection, the boundary boxes and connecting lines between boundary boxes are detected to determine whether the target connected component is a flowchart. 1. A method implemented in a data processing apparatus for recognizing an input document image as a table or a flowchart , the method comprising:(a) detecting a target connected component from the input image which represents candidates of lines of a table or text boxes and connecting lines of a flowchart in the input image;(b) separating the target connected component into a plurality of corners and a plurality of edges that connect the plurality of corners; and(c) based on spatial relationships between the plurality of corners and the plurality of edges, determining whether the target connected component is a table or a flowchart or neither.2. The method of claim 1 , wherein step (a) includes:performing a connected component analysis to identify all connected components in the input image;detecting large connected components that are larger than a threshold size; andmerging some of the large connected components together based on their spatial relationships to form the target connected component.3. The method of claim 1 , wherein step (b) includes:(b1) applying a corner detector operation ...

ON-SHELF IMAGE BASED OUT-OF-STOCK DETECTION

An out-of-stock detection system notifies store management that a product is out of stock. The system captures images of a shelf and determines the position product labels thereon. For each product label, a bounding box is generated based on the position of each product label on the shelf. The system then identifies a product for each product label based on information within each product label and, for each product label, stores a product identified for each bounding box. Accordingly, the system performs an out-of-stock detection process that includes capturing additional image data of the shelf periodically that includes each bounding box, providing a portion of the additional image data for each bounding box to a model trained to determine whether the bounding box contains products, sending a notification for a product determined to be out of stock to a store client device based on output from the model. 1. A method comprising:capturing first image data of a product display area in a store, the product display area including a plurality of product labels, each product label including product information describing a product associated with the product label;determining, from the image data, a position of each of the plurality product labels within the product display area;for each product label, generating a bounding box within the product display area based on the position of each product label within the product display area;identifying a product associated with each product label based on the product information associated with each product label;for each product label, storing an association of the identified product with the bounding box for the product label; and capturing second image data of the product display area that includes each bounding box,', 'providing area portion of the second image data corresponding to each bounding box to a classifier model that is trained to determine whether the bounding box contains products,', 'determining whether one or ...

OBJECT AND DATA POINT TRACKING TO CONTROL SYSTEM IN OPERATION

A computing system obtains image data capturing first and second objects. The system determines, based on user-identified data points, boundaries of the objects and generates a component of a dataset by computing a first data value related to an attribute of a key point in the first image; and computing a second data value related to an attribute of a key point in the first image. The system generates a second component of the dataset, the second component representing updated relative information between the first and second object by generating predicted changes in the first data value and second data value for the second image. The system computes a third data value and a fourth data value related to respective data points in a first and second polygon in the second image. The generating the updated relative information is based on the predicted changes and computed values. 1. A computer-program product tangibly embodied in a non-transitory machine-readable storage medium , the computer-program product including instructions operable to cause a computing system to:obtain data representing a first image of a plurality of images, each of the plurality of images capturing different time points of a system in operation, the system in operation comprising a first object and a second object different from the first object;receive an indication of user-identified data points in a plurality of data points in the data representing the first image, wherein the indication indicates an image coordinate location for each of the user-identified data points in a first coordinate system associated with the first image, and wherein the user-identified data points identify the first object and the second object in the first image;determine, based on a first set of data points of the user-identified data points, a first boundary of the first object in the first image;determine, based on a second set of data points of the user-identified data points, a second boundary of the second ...

APPARATUS AND METHOD FOR EFFICIENTLY STORING RAY TRAVERSAL DATA

Apparatus and method for preventing re-traversal of a prior path on a restart. For example, one embodiment of an apparatus comprises: a ray generator to generate a plurality of rays in a first graphics scene; a bounding volume hierarchy (BVH) generator to construct a BVH comprising a plurality of hierarchically arranged nodes, wherein the BVH comprises a specified number of child nodes at a current BVH level beneath a parent node in the hierarchy; traversal/intersection circuitry to traverse one or more of the rays through the hierarchically arranged nodes of the BVH to form a current traversal path and intersect the one or more rays with primitives contained within the nodes; and traversal tracking circuitry to maintain a path encoding array to store path data related to the current traversal path, the path data comprising an index of a currently traversed child node; wherein the traversal/intersection circuitry is to prevent one or more subsequent rays from re-intersecting primitives from which they originated and/or avoid re-traversing the current traversal path based on the path data in the path encoding array. 1. An apparatus comprising:a ray generator to generate a plurality of rays in a first graphics scene;a bounding volume hierarchy (BVH) generator to construct a BVH comprising a plurality of hierarchically arranged nodes, wherein the BVH comprises a number of child nodes at a current BVH level beneath a parent node in the BVH; andcircuitry to traverse one or more of the rays through the hierarchically arranged nodes of the BVH forming a current traversal path and to intersect the one or more of the rays with primitives contained within the nodes, wherein the circuitry is to process entries from the top of a first data structure comprising entries each associated with a child node at the current BVH level, the entries in the first data structure being ordered from top to bottom based on a sorted distance of each respective child node.2. The apparatus of ...

THREE-DIMENSIONAL BOUNDING BOX FROM TWO-DIMENSIONAL IMAGE AND POINT CLOUD DATA

A three-dimensional bounding box is determined from a two-dimensional image and a point cloud. A feature vector associated with the image and a feature vector associated with the point cloud may be passed through a neural network to determine parameters of the three-dimensional bounding box. Feature vectors associated with each of the points in the point cloud may also be determined and considered to produce estimates of the three-dimensional bounding box on a per-point basis. 1. (canceled)2. A computer-implemented method comprising:receiving sensor data comprising a plurality of measurements of an environment;inputting at least a portion of the sensor data into a machine learned model;determining, as a first feature vector and based at least in part on a first portion of the machine learned model, a first set of values associated with a measurement of the plurality of measurements;determining, as a second feature vector and based at least in part on a second portion of the machine learned model, a second set of values associated with the plurality of measurements;combining, as a combined feature vector, the first feature vector and the second feature vector;inputting the combined feature vector into a third portion of the machine learned model; andreceiving, from the third portion of the machine learned model, information associated with an object represented in the sensor data.3. The computer-implemented method of claim 2 , further comprising:receiving, from an image sensor, image data of the environment;determining a portion of the image data associated with the object;determining, based at least in part on the portion of the image data associated with the object, a subset of the sensor data associated with the portion of the image data;inputting the portion of the image data into a fourth portion of the machine learned model;receiving, from the fourth portion of the machine learned model, an appearance feature vector; andinputting the appearance feature vector ...

IDENTIFICATION OF EMPHASIZED TEXT IN ELECTRONIC DOCUMENTS

To identify emphasized text, bounding boxes are based on clusters resulting from horizontal compression and horizontal morphological dilation. The bounding boxes are processed to determine if any contain words or characters in bold. A bounding box is eliminated based on a comparison of its density and an average density across all bounding boxes. If its density is greater, text elements within the bounding box are evaluated to determine whether the text element is bold. 1. A method of identifying emphasized text , the method comprising:performing horizontal compression on an input image to generate a horizontally compressed image, the input image comprising lines of text, each line of text comprising a plurality of words or characters;performing horizontal morphological dilation on the compressed image to form a horizontally dilated image, the horizontally dilated image comprising clusters, each cluster corresponding to a different one of the lines of text;calculating a bounding box for each cluster, resulting in a plurality of bounding boxes;calculating a first average density, the first average density calculated across all the bounding boxes;for each of the bounding boxes, comparing the first average density to a density of the bounding box; andidentifying a specific bounding box, from among the plurality of bounding boxes, as having a word or character in bold, the identifying based on the comparison of the first average density to the density of the specific bounding box.2. The method of claim 1 , wherein the bounding boxes claim 1 , for calculating the first average density claim 1 , are on the horizontally compressed image.3. The method of claim 1 , wherein:each bounding box includes upper and lower zones, at least one of which contains a fractional part of a word or a character, andnone of the upper zones and none of the lower zones are used in the calculating of the first average density.4. The method of claim 1 , further comprising detecting an underline ...

VEHICLE OPERATION

A computer includes a processor and a memory, the memory storing instructions executable by the processor to determine a first bounding box for a road sign in an image, identify a second bounding box for a physical member, the second bounding box for the physical member extending from the first bounding box for the road sign, and upon identifying the second bounding box for the physical member, actuate one or more vehicle components. 1. A system , comprising a computer including a processor and a memory , the memory storing instructions executable by the processor to:determine a first bounding box for a road sign in an image;identify a second bounding box for a physical member, the second bounding box for the physical member extending from the first bounding box for the road sign; andupon identifying the second bounding box for the physical member, actuate one or more vehicle components.2. The system of claim 1 , wherein the instructions further include instructions to claim 1 , upon identifying no second bounding box extending from the first bounding box claim 1 , determine that the first bounding box indicates an actual road sign upon receiving a message from a server that a plurality of vehicles identified the actual road sign at a location at which the image was collected.3. The system of claim 2 , wherein the instructions further include instructions to determine that the first bounding box indicates the actual road sign when the plurality of vehicles exceeds a threshold.4. The system of claim 2 , wherein the message from the server includes content of the actual road sign claim 2 , and the instructions further include instructions to actuate the one or more components based on the content of the actual road sign.5. The system of claim 1 , wherein the instructions further include instructions to identify a third bounding box for a vehicle and to identify whether the first bounding box is within the third bounding box for the vehicle.6. The system of claim 5 , ...

Method for Training Object Detection Model, Object Detection Method and Related Apparatus

An object detection model training method, object detection method and related apparatus, relate to the field of artificial intelligence technologies such as computer vision, deep learning. An implementation includes: obtaining training sample data including a first remote sensing image and position annotation information of an anchor box of a subject to be detected in the first remote sensing image, where the position annotation information includes angle information of the anchor box relative to a preset direction; obtaining an object feature map of the first remote sensing image based on an object detection model, performing object detection on the subject to be detected based on the object feature map to obtain an object bounding box, and determining loss information between the anchor box and the object bounding box based on the angle information; updating a parameter of the object detection model based on the loss information. 1. A method for training an object detection model , comprising:obtaining training sample data, the training sample data comprising a first remote sensing image and position annotation information of an anchor box of a subject to be detected in the first remote sensing image, wherein the position annotation information comprises angle information of the anchor box relative to a preset direction;obtaining an object feature map of the first remote sensing image based on an object detection model, performing object detection on the subject to be detected based on the object feature map to obtain an object bounding box, and determining loss information between the anchor box and the object bounding box based on the angle information; andupdating a parameter of the object detection model based on the loss information.2. The method according to claim 1 , wherein the angle information is determined at least in part by:obtaining a coordinate sequence of vertices of the subject to be detected in the first remote sensing image, the coordinate ...

Method and system for structural information on-demand

A data system and processes that generate structural characteristics and analytics such as various elevations and heights of a structure. Information produced in on-demand fashion and can be certified. It combines human intelligence with machine intelligence to achieve optimal results. Information produced by the system are significant for many purposes, including Flood Risk Assessment, Flood Insurance Rating, and Flood Impacting Threshold (FIT) determination. The system further generates various derivatives such as Flood Impacting Threshold Score (FITS), precise Flood Risk Ratings (e.g. PrecisionRating,) building conditions and valuation. The system generates such information on-demand by computer vision, artificial intelligence, sensors, image analysis, statistical analysis, and mathematical analysis through a Graphic User Interface (GUI) or machine-to-machine, and Application Programming Interface (API.)

Clustered Palette Compression

Color values may be compressed using a palette based encoder. Clusters of color values may be identified and encoded color values within the cluster with respect to a color value having a predefined characteristic. Clusters that have pixels or samples with constant color value may also be encoded. 1. A method comprising:compressing color values using a palette based encoder;finding clusters of color values and encoding color values within the cluster with respect to a color value having a predefined characteristic; andencoding clusters that have pixels or samples with constant color value.2. The method of wherein said pre-defined characteristic is the sample or pixel with the minimum color value within the cluster.3. The method of including encoding the fact that a color channel is constant.4. The method of including encoding residuals relative to the sample or pixel with the predefined characteristic of the cluster that minimizes cost.5. The method of including refraining from encoding residuals for a cluster with a constant color.6. The method of including using a bit for each channel to encode whether it is constant color or not.7. The method of including providing a group of bits per color channel to indicate how many residual bits are needed per channel.8. The method of wherein finding clusters includes using a K-means clustering algorithm.9. The method of including finding a minimal bounding box around a color.10. The method of including taking a first color from a palette claim 1 , building a cluster around the first color and then claim 1 , if a second color that is different by a threshold from the first color claim 1 , forming a second cluster including the second color claim 1 , and fitting ensuring colors into the first and second clusters.11. One or more non-transitory computer readable media storing instructions executed by a processor to perform a sequence comprising:compressing color values using a palette based encoder;finding clusters of color ...

MOBILE LIDAR PLATFORMS FOR VEHICLE TRACKING

Mobile LiDAR platforms for vehicle tracking are provided. In various embodiments, a time-series of point clouds is received from a LiDAR sensor. Each point cloud of the time series of point clouds is projected onto a plane. The projected point clouds are provided to a convolutional neural network. At least one bounding box of a vehicle within each of the projected point clouds is received from the convolutional neural network. A relative speed of the vehicle is determined from the bounding boxes with the projected point clouds. A reference speed of the LiDAR sensor is determined from the GPS receiver. From the relative speed and the reference speed, an absolute speed of the vehicle is determined. 1. A system comprising:a vehicle-mounted LiDAR sensor;a vehicle-mounted GPS receiver; receiving a time-series of point clouds from the LiDAR sensor;', 'projecting each point cloud of the time series of point clouds onto a plane;', 'providing the projected point clouds to a convolutional neural network;', 'receiving from the convolutional neural network at least one bounding box of a vehicle within each of the projected point clouds;', 'determining from the bounding boxes with the projected point clouds a relative speed of the vehicle;', 'determining from the GPS receiver a reference speed of the LiDAR sensor;', 'from the relative speed and the reference speed, determining an absolute speed of the vehicle., 'a computing node operatively coupled to the LiDAR sensor and GPS receiver, the computing node comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor of the computing node to cause the processor to perform a method comprising2. A method comprising:receiving a time-series of point clouds from the LiDAR sensor;projecting each point cloud of the time series of point clouds onto a plane;providing the projected point clouds to a convolutional neural network;receiving from the ...

INFORMATION PROCESSING APPARATUS RECOGNIZING CERTAIN OBJECT IN CAPTURED IMAGE, AND METHOD FOR CONTROLLING THE SAME

An information processing apparatus includes an image obtaining unit configured to obtain an input image, an extraction unit configured to extract from the input image one or more regions corresponding to one or more objects included in a foreground of the operation surface in accordance with the reflected positional information and positional information of the operation surface in the space, a region specifying unit configured to specify an isolation region, which is not in contact with a boundary line which defines a predetermined closed region in the input image, from among the one or more regions extracted by the extraction unit, and a recognition unit configured to recognize an adjacency state of a predetermined instruction object relative to the operation surface in accordance with the positional information reflected from the portion corresponding to the isolation region as specified by the region specifying unit. 1. An information processing apparatus comprising:an image obtaining unit configured to obtain an input image on which positional information in a space including an operation surface as a portion of a background is reflected;an extraction unit configured to extract one or more regions corresponding to one or more objects included in a foreground of the operation surface from the input image in accordance with the positional information reflected on the input image obtained by the image obtaining unit and positional information of the operation surface in the space;a region specifying unit configured to specify an isolation region which is not in contact with a boundary line which defines a predetermined closed region in the input image from among the one or more regions extracted by the extraction unit; anda recognition unit configured to recognizes an adjacency state of a predetermined instruction object relative to the operation surface in accordance with the positional information reflected on the isolation region in the input image in a ...

Method and system for generating a multi-dimensional surface model of a geometric structure

A method of constructing a bounding box comprises: acquiring a set of sensed data points; adding, for each sensed data point, at least one calculated data point; and defining a bounding box containing the sensed and calculated data points. A method of identifying voxels in a voxel grid corresponding to a plurality of data points comprises: calculating, for each data point, a distance between it and each voxel; creating a subset of voxels comprising voxels having a distance from one data point that is less than a predetermined distance; creating another subset comprising those voxels that neighbor a voxel in the first subset; computing, for each voxel in the second subset, a distance between it and each voxel in the first subset; and identifying each voxel in the first subset that is a distance away from each voxel in the second subset that exceeds a predetermined distance.

MOVEMENT MONITORING SYSTEM

A monitoring system may include an input port, an output port, and a controller in communication with the input port and the output port. The input port may receive video from an image capturing device. The image capturing device is optionally part of the monitoring system and in some cases includes at least part of the controller. The controller may be configured to receive video via the input port and identify a subject within frames of the video relative to a background within the frames. Further, the controller may be configured to identify dimensions and/or other parameters of the identified subject in frames of the video and determine when the subject is performing a predetermined task. Based on the dimensions and/or other parameters identified or extracted from the video during the predetermined task, the controller may output via the output port assessment information. 1. A monitoring system comprising:an input port for receiving video;an output port; identify a subject within a frame of video relative to a background within the frame;', 'determine when the subject in the video is performing a task;', 'identify a height dimension and a width dimension of the subject in one or more frames of the video during the task; and', 'output via the output port position assessment information relative to the subject during the task based on the height dimension and the width dimension for the subject in one or more frames of the video during the task., 'a controller in communication with the input port and the output port, the controller configured to2. The monitoring system of claim 1 , further comprising:an image capturing device adapted to capture video of the subject and provide the video to the controller via the input port.3. The monitoring system of claim 1 , wherein the controller is configured to:determine extreme-most pixels in two dimensions of the subject to identify the height dimension and the width dimension based on the identified subject; andidentify a ...

METHOD FOR CONVERTING LANDSCAPE VIDEO TO PORTRAIT MOBILE LAYOUT USING A SELECTION INTERFACE

Described herein are systems and methods of converting media dimensions. A device may identify a set of frames from a video in a first orientation as belonging to a scene. The device may receive a selected coordinate on a frame of the set of frames for the scene. The device may identify a first region within the frame including a first feature corresponding to the selected coordinate and a second region within the frame including a second feature. The device may generate a first score for the first feature and a second score for the second feature. The first score may be greater than the second score based on the first feature corresponding to the selected coordinate. The device may crop the frame to include the first region and the second region within a predetermined display area comprising a subset of regions of the frame in a second orientation. 1. A method of converting media dimensions , comprising:identifying, by a temporal analyzer executing on a computing device, a set of frames from a video in a first orientation as belonging to a scene;receiving, by a focus region analyzer executing on the computing device, via a user interface, first and second selected coordinates on a frame of the set of frames identified as belonging to the scene;identifying, by an image analyzer executing on the computing device, a first region within the frame comprising a first feature corresponding to the selected coordinate;identifying, by the image analyzer executing on the computing device, within the frame a second region comprising a second feature;generating, by a cropping calculator executing on the computing device, a first score for the first feature and a second score for the second feature, the first score greater than the second score based on the first feature corresponding to the selected coordinate;determining, by the cropping calculator, that the second score exceeds a threshold;cropping, by an image processor executing on the computing device, responsive to the ...

Efficient Label Insertion and Collision Handling

Techniques are described for efficient label insertion and collision handling. A bounding geometry for a label to be graphically displayed on a display screen as part of an electronic map is determined, wherein the bounding geometry comprises a circle. The bounding geometry is inserted into a grid index, wherein the grid index represents a viewport of the electronic map. Disjoint regions of the grid index intersected by the bounding geometry are identified, wherein each disjoint region represents a different portion of the viewport. For each intersected disjoint region, it is identified whether there is at least one collision between the bounding geometry and one or more existing bounding geometries in the disjoint region; and responsive to identifying whether there is at least one collision in the intersected disjoint region, a target opacity of the label is set. 1. A computer-implemented method comprising:determining a bounding geometry for a label as part of an electronic map;inserting the determined bounding geometry into a grid index, wherein the grid index represents a viewport of the electronic map;identifying disjoint regions of the grid index intersected by the determined bounding geometry, wherein different disjoint regions represent different portions of the viewport; identifying a position of the determined bounding geometry within the intersected disjoint region;', 'identifying positions of one or more existing bounding geometries in the disjoint region; and', 'determining whether the identified position of the determined bounding geometry and the identified positions of at least one of the one or more existing bounding geometries overlap; and, 'for one or more intersected disjoint regionsresponsive to identifying that the determined bounding geometry overlaps at least one of the one or more existing bounding geometries, setting a target opacity of the label.2. The method of claim 1 , wherein the determined bounding geometry overlaps at least one of the ...

Machine learning-based part determinations for computer-aided design (cad) assemblies

A computing system may include an assembly access engine configured to access a computer-aided design (CAD) assembly that digitally represents a product component that includes multiple parts. The computing system may also include a part determination engine configured to determine a recommended part for the CAD assembly, including by providing the CAD assembly as an input to a machine-learning (ML) model trained with assembly structure data of CAD assemblies of a common product type as the CAD assembly, generating a candidate part set through the ML model, filtering the candidate part set based on physical and cost characteristics of the different candidate parts of the candidate part set, and identifying the recommended part from the filtered candidate part set. The part recommendation engine may also be configured to insert the recommended part into the CAD assembly and provide the CAD assembly in support of physical manufacture.

METHOD FOR SEGMENTING AN OBJECT IN AN IMAGE

A computer-implemented method for segmenting an object in at least one image acquired by a camera including computing an edge probabilities image based on the image, said edge probabilities image comprising, for each pixel of the image, the probability that said pixel is an edge, computing a segmentation probabilities image based on the image (IM), said segmentation probabilities image comprising, for each pixel of the image (IM), the probability that said pixel belongs to the object (OBJ), and computing a binary mask of the object based on the edge probabilities image and based on the segmentation probabilities image. 1. A computer-implemented method for segmenting an object in at least one image acquired by a camera , comprising:a) computing an edge probabilities image based on the image, said edge probabilities image including, for each pixel of the image, the probability that said pixel is an edge;b) computing a segmentation probabilities image based on the image, said segmentation probabilities image including, for each pixel of the image, the probability that said pixel belongs to the object; andc) computing a binary mask of the object based on the edge probabilities image and based on the segmentation probabilities image.2. The method according to claim 1 , wherein step c) further comprises sub-steps of:c1) building a graph including a plurality of nodes connected one to the other through n-links, each node representing a pixel of the image acquired by the camera, said graph including also a source, which represents the object, and a sink which represents a background of the image, each node being connected to the source and to the sink through t-links; for each n-link, a link cost which is equal to the probability that there is no edge between corresponding adjacent pixels;', 'for each t-link, a source cost which is equal to the probability that the corresponding pixel belongs to the object, and a sink cost which is equal to the probability that the ...

SIMD Group Formation Techniques for Primitive Testing associated with Ray Intersect Traversal

Disclosed techniques relate to primitive testing associated with ray intersection processing for ray tracing. In some embodiments, shader circuitry executes a first SIMD group that includes a ray intersect instruction for a set of rays. Ray intersect circuitry traverses, in response to the ray intersect instruction, multiple nodes in a spatially organized acceleration data structure (ADS). In response to reaching a node of the ADS that indicates one or more primitives, the apparatus forms a second SIMD group that executes one or more instructions to determine whether a set of rays that have reached the node intersect the one or more primitives. The shader circuitry may execute the first SIMD group to shade one or more primitives that are indicated as intersected based on results of execution of the second SIMD group. Thus, disclosed techniques may use both dedicated ray intersect circuitry and dynamically formed SIMD groups executed by shader processors to detect ray intersection.

Path Guiding for Path-Traced Rendering

A computer-implemented method for generating a mask for a light source in a virtual scene includes determining a bounding box for the scene based on a frustum of a virtual camera and generating a path-traced image of the scene within the bounding box. Light paths emitted by the camera and exiting at the light source are stored, and objects poorly sampled by the light source are removed from the scene. An initial mask for the light source is generated from the density of light paths exiting at that position on the light source. The initial mask is refined by averaging in the light path density at each point on the light source for subsequent images. 1. A computer-implemented method for generating a mask for a light source used in rendering virtual scenes , the computer-implemented method comprising: determining a bounding volume for a scene based on a frustum or spherical aperture of a virtual camera having a virtual camera position in the scene;', 'generating a path-traced first image of a portion of the scene within the bounding box by projecting a plurality of first light paths from the virtual camera position, wherein each first light path of the plurality of first light paths comprises a plurality of vertices;', 'storing a light paths subset of the plurality of first light paths, wherein a stored light path in the light paths subset is a light path that exits at the light source;', 'removing objects from the scene that are sampled less than a predetermined threshold by the first light paths to form a modified virtual scene;', 'generating an initial mask for the light source from the light paths subset such that at least some positions on the light source is assigned an emission probability in the initial mask based on a density of first light paths exiting at that position on the light source, thereby generating a masked light source data structure;', (a) generating a path-traced second image using the masked light source data structure and the modified virtual ...

LIGHT PROBE GENERATION METHOD AND APPARATUS, STORAGE MEDIUM, AND COMPUTER DEVICE

A light probe generation method and apparatus, a storage medium, and a computer device are provided. The method includes: selecting shadow points of a target object in a virtual scene; converting the selected shadow points into a voxelized shadow voxel object; reducing a quantity of vertexes in the shadow voxel object to obtain a shadow voxel object after the reduction; and generating a light probe at a vertex position of the shadow voxel object after the vertex reduction. 1. A light probe generation method , performed by a computer device , the method comprising:selecting shadow points of a target object in a virtual scene;converting the selected shadow points into a voxelized shadow voxel object;reducing a quantity of vertexes in the voxelized shadow voxel object; andgenerating a light probe at a vertex position of a shadow voxel object obtained after vertex reduction.2. The method according to claim 1 , wherein the selecting the shadow points comprises:generating a lattice in a bounding box of the virtual scene; anddetermining, with respect to a light source that projects a light ray from the bounding box to the virtual scene, that the light ray is intersected with points in the lattice and the light ray is intersected with the target object, and selecting the intersected points as the shadow points of the target object.3. The method according to claim 2 , further comprising:emitting a ray in a direction of the light source by using the points in the lattice as reference points; andbased on the ray being intersected with target points in the target object, selecting the target points and corresponding reference points as the shadow points of the target object.4. The method according to claim 2 , further comprising:determining a shadow region formed by a virtual object in the light ray;generating a polyhedron enclosing the shadow region; andthe converting the selected shadow points comprises:converting the polyhedron enclosing the shadow region into the voxelized ...

IMAGE PROCESSING APPARATUS AND METHOD

There is provided an image processing apparatus and an image processing method that are capable of suppressing an increase in loads when a point cloud is generated from a mesh. Point cloud data is generated by positioning points at intersection points between a surface of a mesh and vectors each including, as a start origin, position coordinates corresponding to a specified resolution. For example, intersection determination is performed between the surface of the mesh and each of the vectors, and in a case where the surface and the vector are determined to intersect each other, the coordinates of the intersection point are calculated. The present disclosure can be applied to an image processing apparatus, electronic equipment, an image processing method, a program, or the like. 1. An image processing apparatus comprising:a point cloud generating section that generates point cloud data by positioning a point at an intersection point between a surface of a mesh and a vector including, as a start origin, position coordinates corresponding to a specified resolution.2. The image processing apparatus according to claim 1 , wherein performs intersection determination between the surface and the vector, and', 'in a case of determining that the surface and the vector intersect each other, calculates coordinates of the intersection point., 'the point cloud generating section'}3. The image processing apparatus according to claim 2 , whereinthe point cloud generating section performs the intersection determination between the surface and the vector in each of positive and negative directions of each of three axial directions perpendicular to one another.4. The image processing apparatus according to claim 3 , wherein claim 3 ,in a case where multiple intersection points have overlapping coordinate values, the point cloud generating section deletes all intersection points included in a group of the intersection points overlapping each other, except any one of the intersection ...

Method for Immediate Boolean Operations Using Geometric Facets

A method for performing Boolean operations using a computer to create geometric models from primary geometric objects and their facets, comprises calculating intersection lines of facets, splitting facets through which the intersection lines pass, determining each facet is obvious or obscure, and regrouping the facets to form one or more geometric objects. This method does not utilize the most popular data structures CSG and B-REP in CAD/CG/Solid Modeling systems, but has the advantages of both CSG and B-REP: easy to implement and flexible. 1. A method that performs immediate Boolean Operations using geometric facets implemented in a computer system and operating with a computer , the method comprising: mapping rendering facets to extended triangles that contain neighbors; searching for the first pair of triangles that hold a start point of an intersection line by detecting whether two minimum bounding boxes overlap and performing edge-triangle intersection; extending the intersection lines until they get closed by searching neighboring triangles; splitting triangles through which intersection lines pass; Regrouping facets , deleting , and reserving triangles according to Boolean operation types; and mapping extended triangles to rendering facets.2. The method of wherein any Boolean operations claim 1 , including combination claim 1 , intersection claim 1 , exclusion claim 1 , difference claim 1 , and division claim 1 , use rendering facets of the geometric objects to create new geometric objects and the resultants are immediately mapped to rendering triangles without the data structure CSG and B-REP.3. The method of wherein any Boolean operations claim 1 , including combination claim 1 , intersection claim 1 , exclusion claim 1 , difference claim 1 , and division claim 1 , use rendering facets of the geometric objects to create new geometric objects and the resultants are immediately mapped to rendering triangles with the data structure CSG and B-REP.4. The method ...

Hybrid Hierarchy of Bounding and Grid Structures for Ray Tracing

Methods and ray tracing units are provided for performing intersection testing for use in rendering an image of a 3D scene. A hierarchical acceleration structure may be traversed by: traversing one or more upper levels of nodes of the hierarchical acceleration structure according to a first traversal technique, the first traversal technique being a depth-first traversal technique; and traversing one or more lower levels of nodes of the hierarchical acceleration structure according to a second traversal technique, the second traversal technique not being a depth-first traversal technique. Results of traversing the hierarchical acceleration structure are used for rendering the image of the 3D scene. The upper levels of the acceleration structure may be defined according to a spatial subdivision structure, whereas the lower levels of the acceleration structure may be defined according to a bounding volume structure. 1. A computer-implemented method of rendering an image of a 3D scene using a ray tracing system , the method comprising: traversing one or more upper levels of nodes of the hierarchical acceleration structure according to a first traversal technique, said first traversal technique being a depth-first traversal technique; and', 'traversing one or more lower levels of nodes of the hierarchical acceleration structure according to a second traversal technique, said second traversal technique not being a depth-first traversal technique; and, 'performing intersection testing comprising traversing a hierarchical acceleration structure byusing results of said traversing the hierarchical acceleration structure to render the image of the 3D scene.2. The method of wherein the second traversal technique is based on a breadth-first traversal technique claim 1 , wherein intersection testing of nodes with rays is scheduled based on availability of node data and ray data.3. (canceled)4. The method of wherein said one or more upper levels of nodes are at the top of the ...

Hybrid Hierarchy of Bounding and Grid Structures for Ray Tracing

Methods and ray tracing units are provided for performing intersection testing for use in rendering an image of a 3-D scene. A hierarchical acceleration structure may be traversed by: traversing one or more upper levels of nodes of the hierarchical acceleration structure according to a first traversal technique, the first traversal technique being a depth-first traversal technique; and traversing one or more lower levels of nodes of the hierarchical acceleration structure according to a second traversal technique, the second traversal technique not being a depth-first traversal technique. Results of traversing the hierarchical acceleration structure are used for rendering the image of the 3-D scene. The upper levels of the acceleration structure may be defined according to a spatial subdivision structure, whereas the lower levels of the acceleration structure may be defined according to a bounding volume structure. 1. A computer-implemented method of generating a hierarchical acceleration structure and using the hierarchical acceleration structure for intersection testing as part of rendering an image of a 3D scene using a ray tracing system , the method comprising:receiving primitive data for primitives located in the 3D scene;determining nodes of the hierarchical acceleration structure based on the received primitive data, wherein one or more upper levels of nodes of the hierarchical acceleration structure are defined according to a spatial subdivision structure, and wherein one or more lower levels of nodes of the hierarchical acceleration structure are defined according to a bounding volume structure;storing the hierarchical acceleration structure for use in intersection testing; andperforming intersection testing, using the hierarchical acceleration structure, as part of rendering an image of the 3D scene.2. The method of wherein said determining nodes of the hierarchical acceleration structure comprises identifying which primitives are present within ...

METHOD, APPARATUS AND SYSTEM FOR A GRAPHICAL USER INTERFACE

Computer-implemented method of positioning a set of objects on a user interface on a surface of a three-dimensional structure, each member of the set of the objects having a size in two dimensions. Objects are assigned to a respective one of a plurality of subsets of objects based on object characteristics. Each subset is assigned to a respective two-dimensional portion of predetermined size. For each subset having an unpositioned object, object occupancy in the portion is identified and unpositioned objects are allocated to a position within the portion on the basis of object dimension and unoccupied space. The two-dimensional portions are positioned based on a specification and each object position is mapped within the portions to a three-dimensional coordinate. Responsive to input, positioning of the plurality of two-dimensional portions is modified, and each allocated object position is mapped to a new three-dimensional coordinate within the modified portion positions. 1. A computer-implemented method of positioning a set of objects for display on a user interface , the user interface being generated on a user terminal that is configured to present the objects on a surface of a three-dimensional structure , each member of the set of the objects having a size in at least two dimensions , at least some of which dimensions vary between the objects , the method comprising:assigning the set of objects to a two-dimensional portion having a predetermined size;assigning each of the objects to one of a plurality of subsets of objects on the basis of a characteristic of the objects, wherein said characteristic of the objects is distinct from the size of the objects in any dimension and each subset of objects corresponds to a respective band of values for said characteristic of the objects;for at least one of the subsets of objects, performing a positioning process comprising:identifying object occupancy in the portion;allocating the objects in the subset of objects to ...

CENTIMETER HUMAN SKELETON POSE ESTIMATION

A method, apparatus and system for human skeleton pose estimation includes synchronously capturing images of a human moving through an area from a plurality of different points of view, for each of the plurality of captured images, determining a bounding box that bounds the human in the captured image and identifying pixel locations of the bounding box in the image, for each of the plurality of captured images, determining 2D and single-view 3D skeletons from the pixel locations of the bounding box, determining a first, multi-view 3D skeleton using a combination of the 2D and single-view 3D skeletons, and optimizing the first, multi-view 3D skeleton to determine a final 3D skeleton pose for the human. The method, apparatus and system can further include illuminating the area with structured light during the capturing of the images of the human moving through the area. 1. A method for human skeleton pose estimation , comprising:synchronously capturing images of a human moving through an area from a plurality of different points of view;for each of the plurality of captured images, determining a bounding box that bounds the human in the captured image and identifying pixel locations of the bounding box in the image;for each of the plurality of captured images, determining at least one of a 2D skeleton and a single-view 3D skeleton from the identified pixel locations of the bounding box;determining a first, multi-view 3D skeleton using a combination of the at least one of the 2D skeleton and the single view 3D skeleton determined for each of the plurality of captured images; andoptimizing the first, multi-view 3D skeleton to determine a final 3D skeleton pose estimation for the human by applying visibility reasoning techniques to at least some of the 2D skeletons and the single-view 3D skeletons determined for the plurality of captured images.2. The method of claim 1 , comprising:capturing the images of the human moving through the area during an illumination of the ...

AUTOMATED ANNOTATION TECHNIQUES

An annotation system provides various tools for facilitating training data annotation. The annotation tools include a bidirectional annotation model that generates annotations for an image sequence based on both forward information and backward information in an image sequence. The annotation system also facilitates annotation processes by automatically suggesting annotations to the human operator based on a set of annotation predictions and locations of interactions of the human operator on the image. This way, the annotation system provides an accelerated way to generate high-quality annotations that take into account input from a human operator by using the predictions as a guide when it appears that an estimated annotation is consistent with the judgement of the human operator. The annotation system also updates annotations for an overlapping set of objects based on input from human operators. 1. A method of generating annotations for objects in a series of sensor images , comprising:obtaining the sequence of sensor images from one or more sensors, the sequence of sensor images including sensor measurements that sense a surrounding environment of the one or more sensors; generating a set of forward estimations for the sequence of sensor images by propagating a first sensor image of the sequence through the bidirectional machine-learned model in a forward direction,', 'generating a set of backward estimations for the sequence of sensor images by propagating a last sensor image of the sequence through the bidirectional machine-learned model in a backward direction, and', 'combining the set of forward estimations and the set of backward estimations to generate the set of annotations; and, 'generating a set of annotations that indicate locations of the objects in the sequence of sensor images, the set of annotations generated by applying a bidirectional machine-learned model to the sequence of sensor images, comprisingtraining a computer model using the annotated ...

POSE ESTIMATION

A computing system can crop an image based on a width, height and location of a first vehicle in the image. The computing system can estimate a pose of the first vehicle based on inputting the cropped image and the width, height and location of the first vehicle into a deep neural network. The computing system can then operate a second vehicle based on the estimated pose. 1. A method , comprising:cropping an image based on a width, height and center of a first vehicle in the image to determine an image patch;estimating a 3D pose of the first vehicle based on inputting the image patch and the width, height and center of the first vehicle into a deep neural network; andoperating a second vehicle based on the estimated 3D pose.2. The method of claim 1 , wherein the estimated 3D pose includes an estimated 3D position claim 1 , an estimated roll claim 1 , an estimated pitch and an estimated yaw of the first vehicle with respect to a 3D coordinate system.3. The method of claim 1 , further comprising determining the width claim 1 , height and center of the first vehicle image patch based on determining objects in the image based on segmenting the image.4. The method of claim 3 , further comprising determining the width claim 3 , height and center of the first vehicle based on determining a rectangular bounding box in the segmented image.5. The method of claim 4 , further comprising determining the image patch based on cropping and resizing image data from the rectangular bounding box to fit an empirically determined height and width.6. The method of claim 1 , wherein the deep neural network includes a plurality of convolutional neural network layers to process the cropped image claim 1 , a first plurality of fully-connected neural network layers to process the height claim 1 , width and location of the first vehicle and a second plurality of fully-connected neural network layers to combine output from the convolutional neural network layers and the first fully-connected ...

Rendering of Graphic Objects With Pattern Paint Using A Graphics Processing Unit

A technique and computing device to render graphic objects associated with pattern paint in a digital medium environment are disclosed. The computing device comprises a graphics processing unit. The graphics processing unit associates a first surface to a framebuffer of the computing device, renders pattern cells at the first surface, and associates a second surface to the framebuffer of the computing device. The pattern cells correspond to pattern paints, and the pattern cells are rendered at the first surface independent of the sequential order of the pattern cells for rendering pattern paint. The graphics processing unit renders graphic objects associated with the pattern paints, and the graphic objects are rendered at the second surface in sequential order for rendering pattern paint. 1. In a digital medium environment to render graphic objects associated with pattern paint , a method implemented by a computing device utilizing a graphics processing unit , the method comprising:associating, at the graphics processing unit, a first surface to a framebuffer;rendering, at the graphics processing unit, a plurality of pattern cells at the first surface, the plurality of pattern cells associated with a plurality of pattern paints, the plurality of pattern cells being rendered at the first surface independent of their sequential order for rendering pattern paint;associating, at the graphics processing unit, a second surface to the framebuffer; andrendering, at the graphics processing unit, a plurality of graphic objects associated with the plurality of pattern paints, the plurality of graphic objects being rendered at the second surface in sequential order for rendering pattern paint.2. The method as described in claim 1 , wherein rendering the plurality of pattern cells at the first surface includes rendering at least two pattern cells of the plurality of pattern cells at the first surface in parallel.3. The method as described in claim 1 , further comprising: ...

Point Cloud Preprocessing and Rendering

The invention concerns a computer-implemented preprocessing method for dividing a point cloud in cells, whereby for each cell a set of levels of detail is created, each level of detail comprising a set of points and a minimal distance; a data structure for storing a preprocessed point cloud; a computer-implemented rendering method for creating a pixelated image based on a stored preprocessed point cloud and a view frustum; and a viewing method involving a computing system to select a view frustum with a user interface and to present a pixelated image of a point cloud on a screen. 1. Computer-implemented preprocessing method to preprocess a large point cloud , the preprocessing method comprising the steps of:providing a large point cloud, the point cloud comprising points, each of the points at least comprising coordinates which define a location in space;recursively subdividing the point cloud in cells with slice planes, whereby subdivision of a given cell is stopped once the cell contains at most a predetermined amount of points, and whereby each slice plane preferably splits the number of points in a cell approximately in half while preserving locality; subdividing the points at the former finer level of detail of the cell into one or multiple groups;', 'collapsing the points within a group of points to a collapsed point, the collapsed point part of the points at the new coarser level of detail of the cell;, 'for each cell containing at most the predetermined amount of points, creating a set of discrete levels of detail for the cell by recursively constructing a new coarser level of detail from a former finer level of detail, the construction of a new coarser level of detail comprising the steps of geometrical information on the cell subdivision of the point cloud, the geometrical information sufficient to determine a bounding box for each cell;', 'for each cell, the number of levels of detail for the cell, and for each level of detail of the cell the points at ...

Building Recognition via Object Detection and Geospatial Intelligence

A method that enables building recognition via object detection and geospatial intelligence. The method includes detecting an object in an input image, wherein the object is located within a bounding area and extracting a physical location and a camera pose associated with the input image. The method also includes projecting a location coordinate of at least one entity onto the input image to obtain a projected point for the at least one entity, wherein the at least one entity is selected according to the physical location and camera pose associated with the input image and determining a nearest projected point to the bounding area. Finally, the method includes labeling the object within the bounding area with an entity name of the at least one entity associated with the nearest projected point. 1. A method , comprising:detecting an object in an input image, wherein the object is located within a bounding area;extracting a physical location and a camera pose associated with the input image;projecting a location coordinate of at least one entity onto the input image to obtain a projected point for the at least one entity, wherein the at least one entity is selected according to the physical location and camera pose associated with the input image;determining that the projected point is a nearest projected point to the bounding area; andlabeling the object within the bounding area with an entity name of the at least one entity associated with the nearest projected point.2. The method of claim 1 , wherein the object is a structure.3. The method of claim 1 , wherein detecting the object comprises obtaining one or more candidate regions and classifying each proposed candidate region as containing an object or being a background region.4. The method of claim 1 , wherein the bounding area is defined by a location on an image plane and a size of the bounding area.5. The method of claim 1 , wherein the at least one entity is within a predetermined radius of the physical ...

Footstep planning method, robot and computer-readable storage medium

A footstep planning method includes: obtaining a number of depth images of an environment in a walking direction of a legged robot; creating a three-dimensional model of the environment based on the depth images; determining at least one even region from the three-dimensional model of the environment; and selecting one or more of the at least one even region as one or more candidate footstep locations for the legged robot to step on.

FINGERPRINT PREVIEW QUALITY AND SEGMENTATION

A ridge flow based fingerprint image quality determination can be achieved independent of image resolution, can be processed in real-time and includes segmentation, such as fingertip segmentation, therefore providing image quality assessment for individual fingertips within a four finger flat, dual thumb, or whole hand image. A fingerprint quality module receives from one or more scan devices ridge-flow--containing imagery which can then be assessed for one or more of quality, handedness, historical information analysis and the assignment of bounding boxes. 1. An image quality determination method for an input image of any resolution comprising:determining ridge flow for a received image;establishing a ridge flow array; andidentifying one or more regions of the image.258.-. (canceled) This application is a Continuation of U.S. patent application Ser. No. 13/848,908, filed Mar. 22, 2013, which is a Continuation of U.S. patent application Ser. No. 13/047,568, filed Mar. 14, 2011, now. U.S. Pat. No. 8,452,060, which is a Divisional of U.S. patent application Ser. No. 11/426,114, filed Jun. 23, 2006, now U.S. Pat. No. 7,936,907, which claims the benefit of and priority under 35 U.S.C. §119(e) to U.S. patent application Ser. No. 60/795,215, filed Apr. 26, 2006, entitled “Real Time Fingerprint Quality and Segmentation,” each of which are incorporated herein by reference in their entirety.1. Field of the InventionThis invention relates generally to quality determination and segmentation. More specifically, an exemplary aspect of the invention relates to quality and segmentation determination of biometric image information, such as fingerprint, handprint or footprint image information.Livescan type devices provide preview modes of operation that serve as an aid to the operator for finger placement and image fidelity. Because of processing power, timing restrictions and bandwidth requirements, these devices and their associated quality metrics are unable to provide output ...

ARC DETECTION SYSTEM AND METHOD

A method of detecting arcing between two electrical conductors the method including generating an image of the electrical conductors; determining a region of interest in the image where arcing between the electrical conductors is likely to occur, isolating the region of interest processing arcing candidates captured in the image and falling within the region of interest to screen the arcing candidates to reject artefacts likely to have been caused by factors other than arcing, and validating that a selected arcing candidate is an arc by further processing the selected arcing candidate. 119.-. (canceled)21. The method of which includes isolating the region of interest by generating a bounding polygon surrounding a likely contact region between the electrical conductors.22. The method of which includes determining the first sub-region and the second sub-region from the polygon.23. The method of which includes determining the first sub-region by expanding the bounding polygon so that the first sub-region of interest is a first claim 22 , expanded polygon surrounding the bounding polygon and determining the second sub-region by expanding the first claim 22 , expanded polygon to form a second claim 22 , further expanded polygon surrounding the first claim 22 , expanded polygon.24. The method in which processing the arcing candidates to screen the arcing candidates comprises determining if the number of image components of the image claim 20 , having the predetermined characteristic in the first sub-region of the image claim 20 , exceeds the pre-defined threshold and determining if the number of image components of the image claim 20 , having the same characteristic claim 20 , is present in the second sub-region of the image.25. The method of wherein the image components are pixels and wherein the characteristic is associated with intensity.26. The method of which includes using a masking technique to determine if the image component constitutes a centre of the arcing ...

SELF-CALIBRATION OF A STATIC CAMERA FROM VEHICLE INFORMATION

A method to automatically calibrate a static camera in a vehicle is provided. The method may include receiving a captured image file. The method may also include detecting a plurality of vehicles in the captured image file. The method may further include determining a 2D projected shape, size, location, direction of travel, and a plurality of features for each vehicle at various locations in the captured image file. The method may additionally include inferring a plurality of geometry scenes associated with the captured image file, whereby the plurality of geometry scenes is inferred based on the determined 2D projected shape, size, location, direction of travel, and the plurality of features of each vehicle within the detected plurality of vehicles as projected onto the captured image file. The method may include calibrating the static camera based on the inferred plurality of geometry scenes associated with the captured image file. 1. A processor-implemented method to automatically calibrate a static camera in a vehicle , the method comprising:receiving, by a processor, a captured image file;detecting a plurality of vehicles in the captured image file;determining a 2D projected shape, size, location, direction of travel, and a plurality of features for each vehicle within the detected plurality vehicles at various locations in the captured image file;inferring a plurality of geometry scenes associated with the captured image file, wherein the plurality of geometry scenes is inferred based on the determined 2D projected shape, size, location, direction of travel, and the plurality of features of each vehicle within the detected plurality of vehicles as projected onto the captured image file; andcalibrating the static camera based on the inferred plurality of geometry scenes associated with the captured image file.2. The method of claim 1 , further comprising:estimating a size of a plurality of objects in the captured image file; andestimating a speed of the ...

IMAGE PROCESSING METHOD, IMAGE PROCESSING DEVICE AND IMAGE PROCESSING PROGRAM

A system includes circuitry that receives a user input, which may be a gesture. The system smoothes a trajectory of the user input to generate a smoothed trajectory, when the user input is a first gesture that traces the trajectory over an image. The system also applies an effect to the image to cause a target image to be enlarged and displayed on a display screen and to cause the target image to move based on the smoothed trajectory. The target image is a subset of the image. 1. A system comprising: receive a user input, the user input being a gesture,', 'smooth a trajectory of the user input to generate a smoothed trajectory, when the user input is a first gesture tracing the trajectory over an image, and', 'apply an effect to the image to cause a target image to be enlarged and displayed on a display screen and to cause the target image to move based on the smoothed trajectory, the target image being a subset of the image., 'circuitry configured to'}2. The system according to claim 1 , wherein the circuitry is further configured to apply an effect to the image to cause the target image to zoom based on a second gesture operating on the image as the user input.3. The system according to claim 1 , wherein the circuitry causes the target image to move along the smoothed trajectory at a speed corresponding to a speed of the first gesture.4. The system according to claim 1 , wherein the circuitry is further configured to set the target image at a default location on the image based on the user input claim 1 , the target image having a default size.5. The system according to claim 4 , wherein the default location circuitry is further configured to place the target image at a location on the image corresponding to an initial user-touch point on a display surface.6. The system according to claim 5 , wherein the smoothed trajectory starts at the initial user-touch point and ends at a point where the user touch on the display surface ends claim 5 , and the circuitry is ...

TECHNIQUES FOR LABELING, REVIEWING AND CORRECTING LABEL PREDICTIONS FOR P&IDS

In example embodiments, techniques are provided for efficiently labeling, reviewing and correcting predictions for P&IDs in image-only formats. To label text boxes in the P&ID, the labeling application executes an OCR algorithm to predict a bounding box around, and machine-readable text within, each text box, and displays these predictions in its user interface. The labeling application provides functionality to receive a user confirmation or correction for each predicted bounding box and predicted machine-readable text. To label symbols in the P&ID, the labeling application receives user input to draw bounding boxes around symbols and assign symbols to classes of equipment. Where there are multiple occurrences of specific symbols, the labeling application provides functionality to duplicate and automatically detect and assign bounding boxes and classes. To label connections in the P&ID, the labeling application receives user input to define connection points at corresponding symbols and creates connections between the connection points. 1. Method for assigning machine-readable labels to a piping and instrumentation drawing (P&ID) in image-only format , comprising: displaying, in a user interface of a labeling application executing on a computing device, a predicted bounding box and predicted machine-readable text for each of the text boxes, and', 'for at least one of the one or more text boxes, receiving, in the user interface, a correction to the predicted bounding box or predicted machine-readable text and updating the predicted bounding box or predicted machine-readable text;, 'labeling one or more text boxes in the P&ID at least in part by'} displaying in the user interface a bounding box around a given symbol,', 'assigning the given symbol a class of equipment, and', 'automatically detecting one or more other symbols in the P&ID that correspond to the given symbol, placing a bounding box around each of the one or more other symbols, and assigning the class of ...

METHODS AND APPARATUSES FOR DETERMINING FACE IMAGE QUALITY, ELECTRONIC DEVICES, AND COMPUTER STORAGE MEDIA

A method for determining face image quality includes: obtaining pose angle information and/or size information of a face in an image; and obtaining quality information of the face in the image on the basis of the pose angle information and/or the size information of the face. 1. A method for determining face image quality , comprising:obtaining at least one of pose angle information of a face in an image or size information of the face; andobtaining quality information of the face in the image on the basis of at least one of the pose angle information of the face or the size information of the face,wherein the obtaining quality information of the face in the image on the basis of at least one of the pose angle information of the face or the size information of the face comprises:obtaining, according to the pose angle information of the face, a score of a pose angle of the face;obtaining, according to the size information of the face, a score of a face size; andobtaining, according to the score of the pose angle of the face and the score of the face size, a quality score of the face in the image,wherein the obtaining, according to the size information of the face, a score of a face size comprises:obtaining the score of the face size on the basis of at least one of length, width, or area of the face detection bounding box, the area of the face detection bounding box being obtained by a product of the length and width of face detection bounding box,wherein the obtaining the score of the face size on the basis of at least one of length, width, or area of the face detection bounding box comprises:selecting a smaller value min in the length and width of the face detection bounding box; and{'sub': 'rect', 'obtaining, according to the smaller value min, the score Qof the face size by calculation.'}2. The method according to claim 1 , wherein the obtaining the pose angle information of the face in the image comprises:obtaining a face detection bounding box in the image and ...

SYSTEMS AND METHODS FOR UTILIZING MACHINE LEARNING AND OTHER MODELS TO RECONSTRUCT A VEHICLE ACCIDENT SCENE FROM VIDEO

A device may receive, from a first vehicle, video data for video captured of a location associated with an accident for a second vehicle, and may process the video data, with a first model, to generate a sparse point cloud of the location associated with the accident. The device may process the video data, with a second model, to generate depth maps for frames of the video data, and may utilize the depth maps with the sparse point cloud to generate a dense point cloud. The device may process the video data, with a third model, to generate a dense semantic point cloud, and may process the dense semantic point cloud, with a fourth model, to determine a dense semantic overhead view of the location associated with the accident. The device may perform actions based on the dense semantic overhead view. 1. A method comprising:receiving, by a device and from a vehicle device of a first vehicle, video data for video captured of a location associated with an accident associated with a second vehicle;generating, by the device and based on the video data, a sparse point cloud of the location associated with the accident;generating, by the device and based on the video data, depth maps for frames of the video data;utilizing, by the device, the depth maps with the sparse point cloud to generate a dense point cloud;generating, by the device and based on the video data, a dense semantic point cloud;determining, by the device and based on the dense semantic point cloud, a dense semantic overhead view of the location associated with the accident;projecting, by the device, bounding boxes of the second vehicle onto the dense semantic overhead view to provide a trajectory of the second vehicle in the dense semantic overhead view;augmenting, by the device, the dense semantic overhead view with the trajectory of the second vehicle, with additional data to generate a final overhead view; andperforming, by the device, one or more actions based on the final overhead view.2. The method of ...

GENERATING A SHAPE PROFILE FOR A 3D OBJECT

According to an example, a processing device may slice a 3D model along a first direction to generate a plurality of layers in parallel planes defined across a second direction and a third direction, in which each of the plurality of layers is composed of respective polygons representing portions of the 3D model. The plurality of layers may be partitioned into a plurality of stacked boxes containing the respective polygons and for each stacked box of the plurality of stacked boxes, the polygons in the stacked box may be assembled into a superset polygon and the superset polygon may be partitioned into bounding shapes. A shape profile of the 3D object may be generated using the bounding shapes. 1. A computing apparatus comprising:a processing device; access a three dimensional (3D) model of a 3D object;', 'slice the 3D model along a first direction to generate a plurality of layers in parallel planes defined across a second direction and a third direction, wherein the first, second, and third directions are orthogonal to each other, and wherein each of the plurality of layers is composed of respective polygons representing portions of the 3D model;', 'partition the plurality of layers into a plurality of stacked boxes containing the respective polygons;', assemble the polygons in the stacked box into a superset polygon; and', 'partition the superset polygon into bounding shapes; and, 'for each stacked box of the plurality of stacked boxes,'}, 'generate a shape profile of the 3D object using the bounding shapes., 'a non-transitory computer readable storage medium on which are stored instructions that when executed by the processing device cause the processing device to2. The computing apparatus according to claim 1 , wherein the polygons representing portions of the 3D model represent contours of the 3D model claim 1 , and wherein contours represent one of an interior portion and a hole of the 3D model.3. The computing apparatus according to claim 1 , wherein to ...

HIERARCHICAL GENERAL REGISTER FILE (GRF) FOR EXECUTION BLOCK

In an example, an apparatus comprises a plurality of execution units, and a first general register file (GRF) communicatively couple to the plurality of execution units, wherein the first GRF is shared by the plurality of execution units. Other embodiments are also disclosed and claimed. 1. (canceled)2. A general purpose graphics processor , comprising:a general-purpose graphics processing compute block comprising a plurality of processing resources to execute compute instructions, wherein each of the plurality of processing resources comprises a local general register file (GRF) which operates at a first speed;a shared general register file (GRF) communicatively coupled to the plurality of processing resources, wherein the shared GRF operates at a second speed, slower than the first speed.3. The general purpose graphics processor of claim 2 , wherein each of the plurality of processing resources is to retrieve data from the shared GRF when the data is unavailable in the local GRF.4. The general purpose graphics processor of claim 2 , wherein:the local GRF operates at a first power level; andthe shared GRF operates at a second power level, lower than the first power level.5. The general purpose graphics processor of claim 2 , wherein:the local GRF and the shared GRF are separate memory structures.6. The general purpose graphics processor of claim 2 , wherein:the local GRF and the shared GRF are embodied in a single memory structure.7. The general purpose graphics processor of claim 2 , wherein:the shared GRF comprises a virtualized address space.8. The general purpose graphics processor of claim 2 , wherein:each of the plurality of processing resources is to detect a data context switch between two or more of the plurality of processing resources, and in response to the data context switch, to redirect a data inquiry from the local GRF to a remote memory device9. The general purpose graphics processor of claim 2 , further comprising:a shared cache memory ...

ADAPTIVE THREE-DIMENSIONAL SPACE GENERATION METHOD AND SYSTEM

An adaptive three-dimensional space generation system and a method therefor are provided. The adaptive three-dimensional space generation method comprises allowing an adaptive three-dimensional space generation system to determine whether a space is a first-type space or a second-type space depending on structural characteristics of the space based on a plurality of images captured from different locations in the space, allowing the adaptive three-dimensional space generation system to adaptively select an image for texturing the space among the images depending on whether the space is the first-type space or the second-type space, and performing texturing of the space based on the image selected by the adaptive three-dimensional space generation system. 1. An adaptive three-dimensional space generation method comprising:allowing an adaptive three-dimensional space generation system to determine whether a space is a first-type space or a second-type space depending on structural characteristics of the space based on a plurality of images captured from different locations in the space;allowing the adaptive three-dimensional space generation system to adaptively select an image for performing texturing of the space among the images depending on whether the space is the first-type space or the second- type space; andperforming texturing of the space based on the image selected by the adaptive three-dimensional space generation system.2. The adaptive three-dimensional space generation method of claim 1 , wherein said determining whether the space is the first-type space or the second-type space depending on the structural characteristics of the space includes:determining whether the space is the first-type space or the second-type space depending on whether or not at least two criteria are satisfied among a first criterion in which the number of corners of the space is smaller than or equal to a predetermined number, a second criterion in which a ratio of an area of a ...

METHOD AND SYSTEM FOR RESOLVING HEMISPHERE AMBIGUITY USING A POSITION VECTOR

Embodiments resolve hemisphere ambiguity at a system comprising sensors. A hand-held controller of the system emits magnetic fields. Sensors positioned within a headset of the system detect the magnetic fields. A first position and orientation of the hand-held controller is determined within a first hemisphere with respect to the headset based on the magnetic fields. A second position and orientation of the hand-held controller is determined within a second hemisphere, diametrically opposite the first hemisphere, with respect to the headset based on the magnetic fields. A normal vector is determined with respect to the headset, and a position vector identifying a position of the hand-held controller with respect to the headset in the first hemisphere. A dot-product of the normal vector and the position vector is calculated, and the first position and orientation of the hand-held controller is determined to be accurate when a result of the dot-product is positive. 1. A method of resolving hemisphere ambiguity at a system comprising one or more sensors , the method comprising:emitting, at a hand-held controller of the system, one or more magnetic fields;detecting, by one or more sensors positioned within a headset of the system, the one or more magnetic fields;determining a first position and a first orientation of the hand-held controller within a first hemisphere with respect to the headset based on the one or more magnetic fields;determining a second position and a second orientation of the hand-held controller within a second hemisphere with respect to the headset based on the one or more magnetic fields, wherein the second hemisphere is diametrically opposite the first hemisphere with respect to the headset;determining a normal vector with respect to the headset, and a position vector identifying a position of the hand-held controller with respect to the headset in the first hemisphere;calculating a dot-product of the normal vector and the position vector; ...

DEFINING BOUNDARY FOR DETECTED OBJECT

A computer includes a processor and a memory storing instructions executable by the processor to receive data from a sensor specifying a plurality of points, the points including a plurality of first points that describe an object; define a boundary surrounding the first points while minimizing a volume of space that is both (i) contained by the boundary and (ii) identified as unoccupied; and actuate a component with respect to a vehicle based on the boundary. 1. A computer comprising a processor and a memory storing instructions executable by the processor to:receive data from a sensor specifying a plurality of points, the points including a plurality of first points that describe an object;define a boundary surrounding the first points while minimizing a volume of space that is both (i) contained by the boundary and (ii) identified as unoccupied; andactuate a component with respect to a vehicle based on the boundary.2. The computer of claim 1 , wherein the points are described in the sensor data as three-dimensional points claim 1 , and the instructions further include to project the points into two-dimensional horizontal space before defining the boundary.3. The computer of claim 1 , wherein the instructions further include to identify space as unoccupied upon determining that the sensor has an unobstructed view through the space to one of the points.4. The computer of claim 1 , wherein the boundary is a rectangular bounding box.5. The computer of claim 1 , wherein the instructions further include to generate a convex hull surrounding the first points claim 1 , and to define the boundary as a rectangular bounding box based on the convex hull.6. The computer of claim 5 , whereinthe instructions further include to generate a plurality of candidate rectangular bounding boxes based on the convex hull;defining the boundary includes selecting a first candidate rectangular bounding box from the candidate rectangular bounding boxes; andof the candidate rectangular ...

Weakly Supervised Image Segmentation Via Curriculum Learning

Weakly supervised instance segmentation refers to the task of training a system to detect object locations and segment instances of the detected objects, where the training data includes only images and image-level labels. This disclosure includes an enhanced pipeline and enhanced training methods that progressively mine pixel-wise labels, when trained via image-level labels. Four cascaded modules are employed, including: a multi-label classification module, an object detection module, an instance refinement module, and instance segmentation module. The modules share a common backbone. The cascaded pipeline is trained alternatively with a curriculum learning strategy which generalizes image level supervision to pixel level supervision, and a post validation training stage, which runs in the inverse order. In the curriculum learning stage, a proposal refinement sub-module is employed to locate object parts and finding key pixels during classification. 1. A non-transitory computer-readable storage medium having instructions stored thereon for segmenting object instances based on an image segmentation engine that implements a multi-label classification model , an object detection model , an instance refinement model , and an instance segmentation model , which , when executed by a processor of a computing device cause the computing device to perform actions comprising:employing a set of image-level labeled images to supervise a training of the multi-label classification model that generates, based on a first image of the set of image-level labeled images, a first set of object bounding boxes and a first set of weights that corresponds to the first set of object bounding boxes;employing the first set of object bounding boxes and the first set of weights to supervise a training of object detection model that generates, based on the first image, a second set of instance segmentation masks, a second set of object bounding boxes that corresponds to the second set of ...

ALLOCATION OF PRIMITIVES TO PRIMITIVE BLOCKS

An application sends primitives to a graphics processing system so that an image of a 3D scene can be rendered. The primitives are placed into primitive blocks for storage and retrieval from a parameter memory. Rather than simply placing the first primitives into a primitive block until the primitive block is full and then placing further primitives into the next primitive block, multiple primitive blocks can be “open” such that a primitive block allocation module can allocate primitives to one of the open primitive blocks to thereby sort the primitives into primitive blocks according to their spatial positions. By grouping primitives together into primitive blocks in accordance with their spatial positions, the performance of a rasterization module can be improved. For example, in a tile-based rendering system this may mean that fewer primitive blocks need to be fetched by a hidden surface removal module in order to process a tile. 1. A method of processing primitives in a computer graphics processing system in which primitives are allocated to primitive blocks at a primitive block allocation module of a computer graphics processing system , which includes a data store for storing a set of primitive blocks to which primitives can be allocated , wherein a primitive block is configured to store primitive data , the method comprising: (i) comparing an indication of a spatial position of the received primitive with at least one indication of a spatial position of at least one primitive block that is stored in the data store, and', '(ii) allocating the received primitive to a primitive block based on a result of the comparison, such that the received primitive is allocated to a primitive block in accordance with its spatial position; and, 'for each of a plurality of received primitivesprocessing primitive blocks including allocated primitives in the computer graphics processing system.2. The method of claim 1 , wherein said processing primitive blocks comprises ...

GRAPHICS PROCESSING SYSTEMS

Disclosed herein is a bounding box that can be generated for a set of one or more primitive(s) and then passed to a rasteriser circuit for use thereby when generating the graphics fragments to be processed. The bounding box generation integrates a scissor test and allows primitives for which an initial bounding box has zero intersection with a specified scissor box to be discarded, whereas for primitives whose initial bounding box does intersect the scissor box, a new bounding box can be generated for output based on the area of intersection. 1. A method of processing graphics primitives in a graphics processor when generating a render output , the render output comprising a grid of sampling points with respective groups of one or more sampling points of the sampling point grid being associated with respective sub-regions of the render output , the graphics processor comprising:a bounding box generating circuit;a rasteriser circuit that rasterises input primitives to generate graphics fragments to be processed, each graphics fragment having a group of one or more sampling points associated with it; anda renderer circuit that renders fragments generated by the rasteriser circuit,the method comprising, when processing a set of one or more primitive(s) for a render output:obtaining a set of vertex positions for the set of primitive(s);determining a first bounding box using the vertex positions for the set of primitive(s);obtaining a scissor box for the render output;determining an area of intersection between the first bounding box and the scissor box and either:discarding the set of primitive(s), when there is zero intersection between the first bounding box and the scissor box; orwhen there is a non-zero intersection between the first bounding box and the scissor box, generating a second bounding box for the set of primitive(s) based on the area of intersection, and using the second bounding box when rasterising the set of primitive(s);wherein the vertex positions ...

INFORMATION PROCESSING APPARATUS AND INFORMATION PROCESSING METHOD

A plurality of types of arrangement pattern candidates are generated. Each arrangement pattern candidate is an arrangement pattern candidate of a plurality of indices in a physical space used to calculate the position and orientation of the viewpoint. The arrangement pattern candidate enables observation of a predetermined number or more of indices from a position in an area where mixed reality can be experienced in the physical space. Information representing the arrangement pattern of a plurality of indices in the physical space is generated by using the generated types of arrangement pattern candidates. 1. An information processing apparatus comprising:an acquisition unit configured to acquire information of a position, in a physical space, at which a user observes a virtual object;a generation unit configured to generate, on the basis of the information of the position, a location map of indices, the indices being planned to be located in the physical space in order to calculate a position and orientation of a viewpoint of the user; andan output unit configured to output the location map generated by the generation unit.2. The apparatus according to claim 1 , wherein the generation unit generates the location map by generating a plurality of location map candidates of the indices claim 1 , and combining the generated location map candidates.3. The apparatus according to claim 2 , wherein the generation unit generates the location map by deleting at least one index to prevent overlapping of indices in a combined location map obtained by combining the location map candidates.4. The apparatus according to claim 2 , wherein the generation unit generates the location map by moving at least one index to prevent overlapping of indices in a combined location map obtained by combining the location map candidates.5. The apparatus according to claim 2 , wherein the generation unit generates the plurality of location map candidates by generating the location map candidates ...

SKETCH-EFFECT HATCHING

Computer efficiency improved sketch-effect hatching is provided. A hatching tool generates a bounding area associated with an object, within which a plurality of hatching line segment strokes is generated. The strokes are generated by generating center points on equally or near-equally spaced parallel columns within the bounding area that are vertically-offset from each other by randomly-generated stroke lengths. The center points are then randomly horizontally-offset from the columns according to a calculated deviation. A random angle deviation is derived, and end points of the strokes are calculated based on the new center points, stroke lengths, and angle deviations. The strokes are then enhanced by generating overstrokes. Further, the hatching tool maps the strokes to the object, rotates the strokes by a hatching angle, keystone-corrects the strokes, and clips the strokes to fit within the object's boundary. The strokes are output to provide a hand-sketched hatching effect. 1. A computer-implemented method to improve computer-efficiency for generating and applying sketch-effect hatching to an object , comprising:receiving an indication of a selection to apply the sketch-effect hatching to the object;determining a bounding area associated with the object for generating hatching strokes;within the bounding area, generating a plurality of hatching strokes along a plurality of equally or near-equally spaced parallel columns;perturbing each of the plurality of hatching strokes according to a randomly-generated angle deviation; andproviding the object and the plurality of hatching strokes for output to a visual display or a printing device.2. The computer-implemented method of claim 1 , wherein generating the plurality of hatching strokes along the plurality of equally or near-equally spaced parallel columns comprises generating the plurality of hatching strokes claim 1 , wherein each of the plurality of hatching strokes has a randomly-generated stroke length.3. The ...

GRAPHICS PROCESSING SYSTEMS

In a graphics processing system, a bounding volume () representative of the volume of all or part of a scene to be rendered is defined. Then, when rendering an at least partially transparent object () that is within the bounding volume () in the scene, a rendering pass for part or all of the object () is performed in which the object () is rendered as if it were an opaque object. In the rendering pass, for at least one sampling position () on a surface of the object (), the colour to be used to represent the part of the refracted scene that will be visible through the object () at the sampling position () is determined by using a view vector () from a viewpoint position () for the scene to determine a refracted view vector () for the sampling position (), determining the position on the bounding volume () intersected by the refracted view vector (), using the intersection position () to determine a vector () to be used to sample a graphics texture that represents the colour of the surface of the bounding volume () in the scene, and using the determined vector () to sample the graphics texture to determine a colour for the sampling position () to be used to represent the part of the refracted scene that will be visible through the object () at the sampling position () and any other relevant information encoded in one or more channels of the texture. 1. A method of operating a graphics processing system when rendering a scene for output , in which a bounding volume representative of the volume of all or part of the scene to be rendered is defined; the method comprising:when rendering an at least partially transparent object that is within the bounding volume in the scene:performing a rendering pass for some or all of the object in which the object is rendered as if it were an opaque object; andin the rendering pass: using a view vector from a viewpoint position for the scene to determine a refracted view vector for the sampling position;', 'determining the position on ...

ENHANCED OBJECT IDENTIFICATION USING ONE OR MORE NEURAL NETWORKS

Apparatuses, systems, and techniques to identify one or more objects in one or more images. In at least one embodiment, one or more objects are identified in one or more images based, at least in part, on a likelihood that one or more objects is different from other objects in one or more images. 1. A processor , comprising: one or more circuits to use one or more neural networks to identify one or more objects in one or more images based , at least in part , on a likelihood that the one or more objects is different from other objects in the one or more images.2. The processor of claim 1 , wherein the one or more circuits are to:determine probability distributions for a set of properties of a plurality of bounding boxes detected in the one or more images;calculate one or more uncertainty values from the probability distributions; andaggregate the one or more uncertainty values to determine one or more informativeness values for the one or more images.3. The processor of claim 2 , wherein the one or more neural networks are to further:compare the one or more informativeness values with each other;select a subset of the one or more images based at least in part on the comparison of the one or more informativeness values with each other; andtrain a neural network using the subset of the one or more images.4. The processor of claim 2 , wherein one or more circuits are to:determine a Gaussian mixture model (GMM) from the probability distributions and weights corresponding to the probability distributions; anduse the GMM to calculate an aleatoric uncertainty value and epistemic uncertainty value.5. The processor of claim 4 , wherein the one or more circuits are to train the GMM using a mixture density network to regress one or more localization parameters of the GMM using a negative log-likelihood loss function.6. The processor of claim 4 , wherein the one or more circuits are to train the GMM using a mixture density network to regress one or more classification ...

HEATMAP VISUALIZATION OF OBJECT DETECTIONS

An electronic apparatus and method for heatmap visualization of object detections is provided. The electronic apparatus inputs an image frame including an object of interest to Deep Neural Network (DNN) model and extracts an object detection result associated with the object of interest as output of the DNN model for the input image frame. The object detection result includes bounding box coordinates for the object of interest. The electronic apparatus segments the input image frame into a plurality of image portions based on the bounding box coordinates and determines, for each of the plurality of image portions, a weight value indicative of a likelihood that a corresponding image portion belongs to the object of interest. The electronic apparatus generates, based on the determined weight value, a visualization frame which includes heatmap visualization of a region that is included in the image frame and is bounded by the bounding box coordinates. 1. An electronic apparatus , comprising:{'claim-text': ['input a first image frame comprising an object of interest to a Deep Neural Network (DNN) model which is trained for an object detection task;', {'claim-text': 'the extracted object detection result comprising bounding box coordinates for the object of interest;', '#text': 'extract an object detection result associated with the object of interest from an output of the DNN model for the input first image frame,'}, 'segment the input first image frame into a first plurality of image portions based on the bounding box coordinates;', 'determine, for each image portion of the first plurality of image portions, a weight value indicative of a likelihood that a corresponding image portion belongs to the object of interest; and', 'generate, based on the determined weight value for each image portion of the first plurality of image portions, a visualization frame comprising a heatmap visualization of a region which is included in the input first image frame and is bounded by ...

APPARATUS AND METHOD FOR COMPRESSING RAY TRACING ACCELERATION STRUCTURE BUILD DATA

Apparatus and method for compression of acceleration structure build data in a ray tracing implementation. For example, one embodiment of an apparatus comprises: traversal hardware logic to traverse rays through a graphics scene comprising a plurality of primitives; and an acceleration data structure processing unit comprising: a bounding box compressor to compress a set of bounding boxes to generate a plurality of bounding box compression blocks, and an index compressor to compress a set of indices to generate a plurality of index compression blocks, and an acceleration data structure builder for constructing acceleration structures based on bounding box compression blocks and index compression blocks 1. An apparatus comprising:traversal hardware logic to traverse rays through a graphics scene comprising a plurality of primitives; and{'claim-text': ['a bounding box compressor to compress a set of bounding boxes to generate a plurality of bounding box compression blocks,', 'an index compressor to compress a set of indices to generate a plurality of index compression blocks, and', 'an acceleration data structure builder to process the bounding box compression blocks and index compression blocks to generate acceleration data structures'], '#text': 'an acceleration data structure processing unit comprising:'}2. The apparatus of wherein the acceleration data structure comprises a bounding volume hierarchy (BVH) to be used by the traversal hardware logic to traverse the rays through the graphics scene claim 1 , the bounding boxes of the BVH stored as a number of axis-aligned bounding boxes (AABBs) comprising a hierarchical representation of the plurality of primitives.3. The apparatus of wherein the bounding box compressor is to use delta compression to compress a first plurality of the bounding boxes in a first bounding box compression block and the index compressor is to use delta compression to compress a first plurality of the indices in a first index compression ...

Neural Network Foreground Separation for Mixed Reality

A head wearable device, a method, and a system. The head wearable device may include a display, a camera, a convolutional neural network (CNN) processor, and a processor. The CNN processor may be configured to: receive real scene image data from the camera; identify and classify objects in a real scene image; and generate object classification and position data. The processor may be configured to receive the real scene image data; receive the object classification and position data from the CNN processor; perform an image segmentation operation on the real scene image to fill in the objects; generate filled-in object data indicative of filled-in objects; generate a pixel mask; receive virtual scene image data; create mixed reality scene image data; and output the mixed reality scene image data to the display for presentation to a user. 1. A head wearable device , comprising:a display;a camera configured to capture a real scene image; receive real scene image data corresponding to the captured real scene image from the camera;', 'identify objects in the real scene image;', 'classify each object of the identified objects as a foreground object or a background object;', 'generate object classification and position data indicative of classifications and positions of the identified objects; and', 'output the object classification and position data; and, 'at least one convolutional neural network (CNN) processor communicatively coupled to the camera, the at least one CNN processor being at least one of at least one graphics processing unit (GPU) or at least one field-programmable gate array (FPGA), the at least one CNN processor configured to receive the real scene image data corresponding to the captured real scene image from the camera;', 'receive the object classification and position data from the at least one CNN processor;', 'based at least on position data of the object classification and position data, perform at least one image segmentation operation on the real ...

Systems and methods for real-time complex character animations and interactivity

Systems, methods, and non-transitory computer-readable media can identify a virtual deformable geometric model to be animated in a real-time immersive environment. The virtual deformable geometric model comprises a virtual model mesh comprising a plurality of vertices, a plurality of edges, and a plurality of faces. The virtual model mesh is iteratively refined in one or more iterations to generate a refined mesh. Each iteration of the one or more iterations increases the number of vertices, the number of edges, and/or the number of faces. The refined mesh is presented during real-time animation of the virtual deformable geometric model within the real-time immersive environment.

Systems and methods for real-time complex character animations and interactivity

Systems, methods, and non-transitory computer-readable media can receive virtual model information associated with a virtual deformable geometric model. The virtual model information comprises a complex rig comprising a plurality of transforms and a first plurality of vertices defined by a default model, and a simplified rig comprising a second plurality of transforms and a second plurality of vertices. The second plurality of vertices correspond to the first plurality of vertices defined by the default model. The simplified rig and the complex rig are deformed based on an animation to be applied to the virtual deformable geometric model. A set of offset data is calculated. The set of offset data comprises, for each vertex in the first plurality of vertices, an offset between the vertex and a corresponding vertex in the second plurality of vertices.

SYSTEMS AND METHODS FOR REAL-TIME COMPLEX CHARACTER ANIMATIONS AND INTERACTIVITY

Systems, methods, and non-transitory computer-readable media can receive virtual model information associated with a virtual deformable geometric model. The virtual model information comprises a complex rig comprising a plurality of transforms and a first plurality of vertices defined by a default model, and a simplified rig comprising a second plurality of transforms and a second plurality of vertices corresponding to the first plurality of vertices. The simplified rig and the complex rig are deformed based on an animation to be applied to the virtual deformable geometric model. A set of offset data is calculated. The set of offset data comprises, for each vertex in the first plurality of vertices, an offset between the vertex and a corresponding vertex in the second plurality of vertices. A compressed version of the set of offset data is exported to a real-time processing engine for real-time animation of the virtual deformable geometric model. 1. A computer-implemented method comprising: a complex rig comprising a first plurality of transforms and a first plurality of vertices defined by a default model, and', 'a simplified rig comprising a second plurality of transforms and a second plurality of vertices, wherein the second plurality of vertices correspond to the first plurality of vertices defined by the default model;, 'receiving, by a computing system, virtual model information associated with a virtual deformable geometric model, the virtual model information comprising'}deforming, by the computing system, the simplified rig and the complex rig based on an animation to be applied to the virtual deformable geometric model;calculating, by the computing system, a set of offset data, the set of offset data comprising, for each vertex in the first plurality of vertices, an offset between the vertex and a corresponding vertex in the second plurality of vertices; andexporting, by the computing system, a compressed version of the set of offset data to a real-time ...

Methods, Computer Program and Apparatus for an Ordered Traversal of a Subset of Nodes of a Tree Structure and for Determining an Occlusion of a Point along a Ray in a Raytracing Scene

Examples relate to methods, a computer program and an apparatus for an ordered traversal of a subset of nodes of a tree structure and/or for determining an occlusion of a point along a ray in a raytracing scene. The method for the ordered traversal of the subset of nodes of the tree structure comprises obtaining ordering information indicating a desired order of the ordered traversal of the tree structure. The method further comprises selecting a predetermined ordering parameter template from a plurality of predetermined ordering parameter templates based on the ordering information. The method further comprises copying the subset of nodes of the tree structure from a first memory region to a second memory region using a single processor operation of a vector processing processor instruction set such that the subset of nodes of the tree structure is stored within the second memory region in the desired order. The selected predetermined ordering parameter template is used as a parameter for the single processor operation. 1. A method for an ordered traversal of a subset of nodes of a tree structure , the method comprisingobtaining ordering information indicating a desired order of the ordered traversal of the tree structure;selecting a predetermined ordering parameter template from a plurality of predetermined ordering parameter templates based on the ordering information; andcopying the subset of nodes of the tree structure from a first memory region to a second memory region using a single processor operation of a vector processing processor instruction set such that the subset of nodes of the tree structure is stored within the second memory region in the desired order, wherein the selected predetermined ordering parameter template is used as a parameter for the single processor operation.2. The method according to claim 1 , wherein the ordering information indicates a query order index claim 1 , and wherein the predetermined parameter template assigned to the ...

IMAGE-BASED CLASSIFICATION SYSTEM

The present disclosure provides an image-based classification system, comprising an image capturing device, and a processing device connected to the image capturing device. The image capturing device is used for capturing an image of an object. The object has a surface layer and an inner layer. The processing device is configured to use a deep learning model, perform image segmentation on the image of the object, define a surface-layer region and an inner-layer region of the image, and generate classification information. The surface-layer region and the inner-layer region correspond respectively to the surface layer and the inner layer of the object. 1. An image-based classification system , comprising:an image capturing device, adapted for capturing an image of an object, wherein the object has a surface layer and an inner layer; anda processing device, connected to the image capturing device and configured to use a deep learning model, perform image segmentation on the image of the object, define a surface-layer region and an inner-layer region of the image, and generate classification information, wherein the surface-layer region and the inner-layer region correspond respectively to the surface layer and the inner layer of the object.2. The image-based classification system of claim 1 , wherein the processing device is configured to use the deep learning model claim 1 , performs inspection according to the classification information claim 1 , identifies any defect in the inner-layer region claim 1 , and outputs an inspection result.3. The image-based classification system of claim 2 , wherein the deep learning model comprises:a backbone network, adapted for performing feature extraction on an original image of the object and thereby obtaining at least one feature map;a region proposal network (RPN), connected to the backbone network and configured to obtain the feature map from the backbone network and determine at least one region of interest according to the ...

Apparatus and method for optimized tile-based rendering

A virtual reality apparatus and method are described for tile-based rendering. For example, one embodiment of an apparatus comprises: a set of on-chip geometry buffers including a first buffer to store geometry data, and a set of pointer buffers to store pointers to the geometry data; a tile-based immediate mode rendering (TBIMR) module to perform tile-based immediate mode rendering using geometry data and pointers stored within the set of on-chip geometry buffers; spill circuitry to determine when the on-chip geometry buffers are over-subscribed and responsively spill additional geometry data and/or pointers to an off-chip memory; and a prefetcher to start prefetching the geometry data from the off-chip memory as space becomes available within the on-chip geometry buffers, the TBIMR module to perform tile-based immediate mode rendering using the geometry data prefetched from the off-chip memory.

SYSTEM & METHOD FOR GENERATING A STEREO PAIR OF IMAGES OF VIRTUAL OBJECTS

A method and system for generating a stereo pair of images of a 3D virtual space and presenting the images to a user. The system includes a display unit configured to displaying a stereo pair of images to the user and a processor for generating the stereo pair of images. The method of the invention, carried out by the processor includes positioning virtual scenes S1, . . . Sn, . . . SN in the virtual 3D space. The scenes are subjected to a first manipulation and then a left 2D image is obtained by projecting the space onto a viewing surface positioned in the 3D space. The scenes are then subjected to a second manipulation and the right 2D image is obtained by projecting the space onto the viewing surface. 1: A method for generating a stereo pair of images , of one or more scenes , the stereo pair comprising a 2D left view image and a 2D right view image the method comprising:(a) generating a virtual 3D space having an x-axis, a y-axis and a z-axis, the x-axis, the y-axis and the z-axis being mutually orthogonal;{'sub': 1', '1, '(b) positioning a collection of an integer N1 of scenes S, . . . Sn, . . . SN in the virtual 3D space, each of the scenes S, . . . Sn, . . . SN having a first associated predetermined point of rotation An;'}(c) a 2D viewing surface and a viewing ray having vertex or viewpoint in the virtual space, the viewing surface being positioned in the virtual 3D space parallel to the y axis and orthogonal to the z axis, the viewing ray being normal to the viewing surface and the viewing surface and the viewing ray determining a perspective projection of the 3D space onto the viewing surface;(d) rotating each of the scenes Sn in the virtual 3D space about an axis of rotation passing through the predetermined point An, the rotation being through a predetermined angle 8n, where 8n>O for at least one n, and the rotation of Sn in the virtual 3D space being counter-clockwise when viewed from a point on the axis of rotation viewed towards the x-z plane and in ...

METHOD AND SYSTEM FOR DETERMINING COMPLEX SITUATION ACCORDING TO TIME FLOW OF EVENTS OCCURRING IN EACH DEVICE

Provided is a method and system for determining a complex situation according to the time flow of events occurring in each device. In a method for determining a complex situation according to an embodiment of the present invention, a current situation is determined by detecting events occurring in devices and referring to a complex situation determination rule. The complex situation determination rule is a rule in which the current situation is mapped according to the events, which have occurred in the devices, and time intervals between the events. The present invention enables a more reliable determination of the current situation by complexly considering not only the events occurring in the devices but also time intervals between the events. 1. A method for determining a complex situation , the method comprising the steps of:detecting events occurring in devices; anddetermining a current situation by referring to a result of the detecting in the step of detecting and a complex situation analytic rule, andwherein the complex situation analytic rule is a rule which maps a current situation according to the events occurring in the devices and time intervals between the events.2. The method of claim 1 , wherein the complex situation analytic rule comprises a first condition and a first situation which is mapped onto the first condition claim 1 , andwherein the first condition comprises a first detection condition in which a first event is detected in a first device, and a second detection condition in which a second event is detected in a second device before a first time elapses after the first event of the first device.3. The method of claim 2 , wherein the first condition further comprises a third detection condition in which a third event is detected in a third device before a second time elapses after the second event of the second device.4. The method of claim 2 , wherein the first condition further comprises a fourth detection condition in which a fourth event ...

Method for detecting collision between cylindrical collider and convex body in real-time virtual scenario, terminal, and storage medium

A method for detecting a collision between a cylindrical collider and a convex body in a real-time virtual scenario performed at a computer includes: converting a cylindrical collider into a preset polygonal prism concentric to the cylindrical collider; transforming the preset polygonal prism to a local coordinate system of the convex body; obtaining a projection of the cylindrical collider on one or more testing axes according to each testing axis and the location of the preset polygonal prism in the local coordinate system of the convex body, and obtaining a projection of the convex body on each testing axis; and in accordance with a determination that the projections of the cylindrical collider and the convex body intersect with each other on each testing axis, moving the cylindrical collider away from the convex body in the real-time virtual scenario to avoid the collision.

METHOD AND SYSTEM FOR DISPLAYING A LARGE 3D MODEL ON A REMOTE DEVICE

A method of displaying a digital object on a device is disclosed. A set of unique identifiers and associated is downloaded for a set of digital objects within a 3D model. A position and an orientation of a virtual camera is determined relative to the positions of the set of digital objects. A visibility score for each digital object of the set of digital objects is computed based on a position of the digital object relative to the position and the orientation of the virtual camera. A polygon mesh associated with a digital object with a highest visibility score is downloaded based on the polygon mesh being absent. The downloading uses a unique identifier of the set of unique identifiers associated with the polygon mesh. The polygon mesh of the digital object with the highest score is rendered and displayed on a display device. 1. A system comprising:one or more computer processors;one or more computer memories;a set of instructions stored in the one or more computer memories, the set of instructions configuring the one or more computer processors to perform operations comprising:downloading a set of unique identifiers and associated positions for a set of digital objects within a 3D model;determining a position and an orientation of a virtual camera relative to the positions of the set of digital objects;computing a visibility score for each digital object of the set of digital objects based on a position of the digital object relative to the position and the orientation of the virtual camera;{'b': '1', 'downloading a polygon mesh associated with a digital object with a highest visibility score based on the polygon mesh being absent from the one or more computer memories, the downloading using a unique identifier of the set of unique identifiers; and p rendering and displaying the polygon mesh of the digital object with the highest score on a display device.'}2. The system of claim 1 , the downloading of the polygon mesh further including:downloading the polygon mesh ...

METHOD AND APPARATUS OF DETECTING OBJECT OF INTEREST

Disclosed is a method and apparatus of detecting an object of interest, where the apparatus acquires an input image, sets a region of interest (ROI) in the input image, and detects the object of interest from a restoration image, having a resolution greater than a resolution of the input image, corresponding to the ROI. 1. A method of detecting an object of interest , the method comprising:acquiring an input image;setting a region of interest (ROI) in the input image;generating a restoration image corresponding to the ROI, a resolution of the restoration image being greater than a resolution of the input image; anddetecting the object of interest from the restoration image.2. The method of claim 1 , wherein the setting of the ROI comprises:detecting at least one object comprising a lane marking in the input image; andsetting the ROI based on the at least one object.3. The method of claim 1 , wherein the setting of the ROI comprises:detecting a road vanishing point where a lane marking converges based on a segmentation of the lane marking in the input image; andsetting a region corresponding to the road vanishing point as the ROI.4. The method of claim 3 , wherein the region corresponding to the road vanishing point comprises a region within a distance from the road vanishing point claim 3 , and the distance is an inverse function of a number of lanes on a road.5. The method of claim 1 , wherein the setting of the ROI comprises:calculating a reliability score of at least one object in the input image; andsetting the ROI based on the reliability score.6. The method of claim 5 , wherein the calculating of the reliability score comprises calculating the reliability score of the at least one object in the input image by detecting a bounding box based on deep learning or image processing.7. The method of claim 6 , wherein the setting of the ROI comprises setting a region of the bounding box as the ROI based on a comparison of the reliability score to a threshold value.8. ...

METHOD FOR THE COMPUTER ANIMATION OF CAPTURED IMAGES

A method for the computer animation of captured images is provided. The computer animation of the captured image may be operatively rigged via touchscreen input or synchronized with video and audio input. During the rigging stage, the user may locate joints and pivot points and/or isolated body parts via a touchscreen. Animation may be based on, in part, motion capture from video input devices, and by using computer vision and machine learning to predict the location of the pivot points and associated body parts. 1. A method of computer animating captured images , comprising:obtaining a first set of rules that define one or more first nodes as a function of two or more first body parts of one or more animatable captured images;obtaining a timed data file of a moving body having one or more tracked coordinates of two or more second body parts of said moving body;evaluating the one or more tracked coordinates against the first set of rules to determine a respective tracked coordinate for each first node; andapplying synchronized movement to each first node by mirroring each respective tracked coordinate to produce animation control of the one or more animatable captured images.2. The method of claim 1 , wherein the timed data file of a moving body is a video output of a user.3. The method of claim 1 , wherein the first set of rules comprises one or more touchscreen gestures defining each first node.4. The method of claim 1 , wherein the evaluation of the one or more tracked coordinates against the first set of rules comprises inverse kinematic logic functionality is configured to predict motion of some first nodes even if their respective tracked coordinate is not directly tracked.5. The method of claim 4 , wherein the inverse kinematic logic functionality infers a continued motion at a continuous velocity.6. The method of claim 1 , the first set of rules comprise defining a relationship between two adjacent first nodes as a vector claim 1 , wherein each vector ...

3D User Interface - Non-native Stereoscopic Image Conversion

Systems and methods for displaying a three-dimensional (3D) workspace, including a 3D internet browser, in addition to a traditional two-dimensional (2D) workspace and for browsing the internet in a 3D/virtual reality workspace and transforming and/or upconverting objects and/or visual media from the 2D workspace and/or 2D webpages to the 3D workspace as 3D objects and/or stereoscopic output for display in the 3D workspace. 1. A non-transitory computer readable memory medium storing programming instructions executable by a processor of a three-dimensional (3D) display system to:detect, within a web browser, stereoscopic content in a two-dimensional (2D) webpage structure, wherein a 2D webpage associated with the 2D webpage structure is displayed within the web browser via at least one display of the 3D display system, and wherein the stereoscopic content is associated with at least a portion of content of the 2D webpage;convert the detected stereoscopic content to a 3D object; andrender, via the at least one display and within the web browser, the 3D object.2. The non-transitory computer readable memory medium of claim 1 ,wherein the stereoscopic content is included in the 2D webpage structure via a non-natively stereoscopic format.3. The non-transitory computer readable memory medium of claim 2 , a side-by-side format;', 'a top-bottom format;', 'an interleaved format; or', 'an anaglyph., 'wherein the non-natively stereoscopic format comprises one of4. The non-transitory computer readable memory medium of claim 1 , searching for at least one of images and videos; or', 'scheduling a re-search process for dynamic changes to the 2D webpage, wherein the dynamic changes occur after the 2D webpage loads., 'wherein, to detect the stereoscopic content, the programming instructions are further executable to perform at least one of5. The non-transitory computer readable memory medium of claim 4 , 'analyze search results, including searching the search results for at least one ...

SYSTEM AND METHOD FOR COMPRESSING AND DECOMPRESSING TIME-VARYING SURFACE DATA OF A 3-DIMENSIONAL OBJECT USING A VIDEO CODEC

A processor implemented method for compressing time-varying surface data of a 3 dimensional object in a global digital space having frames, using a video encoder that supports a video data compression algorithm, the video encoder being coupled to a transmitter. The method includes the steps of (i) decomposing the time-varying surface data into at least one surface representation that is encoded in an oriented bounding box, (ii) transforming the oriented bounding box into a canonical camera representation for each frame to obtain canonical coordinates for the at least one surface representation, (iii) converting each of the at least one surface representation into at least one bounding box video pair that includes a grayscale video representing depth, and a color video and (iv) tiling the at least one bounding box video pair for each frame to produce a tiled bounding box video. 1. A processor implemented method for compressing time-varying surface data of a 3 dimensional object in a global digital space having frames , using a video encoder that supports a video data compression algorithm , the video encoder being coupled to a transmitter , the method comprising:decomposing the time-varying surface data into at least one surface representation that is encoded in an oriented bounding box, wherein the at least one surface representation is parameterized in 2 dimensions as a depth map with color data, and the oriented bounding box is parameterized to define a location and an orientation in the global digital space;transforming the oriented bounding box into a canonical camera representation for each frame to obtain canonical coordinates for the at least one surface representation;converting each of the at least one surface representation into at least one bounding box video pair that comprises a grayscale video representing depth, and a color video, wherein the depth map is encoded by the video encoder as the grayscale video, and wherein the color video represents ...

SYSTEM AND METHOD FOR 3D BLOB CLASSIFICATION AND TRANSMISSION

Embodiments described herein provide an apparatus comprising a processor to project and accumulate three-dimensional (3D) point data from a blob onto one of a floor plane or a ceiling plane, construct a histogram of the 3D point data, determine a center of mass of the point data from the blob, obtain height data for the 3D point data, and calculate dimensions for a bounding box to surround the blob. Other embodiments may be described and claimed. 1. A method , comprising:projecting and accumulating three-dimensional (3D) point data from a blob onto one of a floor plane or a ceiling plane;constructing a histogram of the 3D point data;determining a center of mass of the point data from the blob;obtaining height data for the 3D point data; andcalculating dimensions for a bounding box to surround the blob.2. The method of claim 1 , further comprising:encoding the bounding box with coordinates for two opposing corners of the bounding box.3. The method of claim 1 , further comprising:implementing a classification algorithm to identify one or more blobs in a point cloud data set.4. The method of claim 3 , wherein the classification algorithm comprises a K-means classification algorithm.5. The method of claim 1 , further comprising:calculating a two-dimensional (2D) distance transform for the blob to generate peaks in coordinates for data in the blob; andsurrounding the peaks with a 2D rectangle defined by a diameter of the blob.6. The method of claim 5 , further comprising:merging the 2D rectangle with the height data to define the bounding box.7. A non-transitory machine readable medium storing instructions which claim 5 , when executed by one or more processors claim 5 , cause the one or more processors to perform operations comprising:projecting and accumulating three-dimensional (3D) point data from a blob onto one of a floor plane or a ceiling plane;constructing a histogram of the 3D point data;determining a center of mass of the point data from the blob;obtaining ...

Ai-based physical function assessment system

Various embodiments of devices, systems, and methods for providing AI-based physical function assessment recordings and assessment performance analytics for a subject are described. A series of video frames are obtained that include the subject. Computer vision techniques that use artificial neural networks may be applied to the video frames to: detect a Person of Interest (POI) and an Object of Interest (OOI) in the video frames; track movement of the POI and the location of the OOI in subsequent video frames; detect body key points; and detect postures and posture transitions of the POI. Physical function indicators may be calculated and function analytics provided based on the assessment.

METHODS AND APPARATUS TO DETECT EFFECTIVE TILING AREA AND FILL TILES EFFICIENTLY

The disclosure is directed towards methods and apparatus to detect effective tiling area and fill tiles efficiently. The method improves efficiency by not filling tiles within an inner box in a shape having a large unfilled area. One example method includes detecting an inner box, determining whether the detected inner box is big enough for pre-clipping, and confirming that the outer clip path contains the inner box. When filling tiles into a bounding rectangle tiling area, it is determined if a particular tile (or tile(s)) falls into an inner box or not, and if the tile falls in the inner box, that particular tile is not filled. According to one embodiment, the inner box is an internal rectangle that contains a maximum area in which it is unnecessary to fill tiles. 1. A method comprising:detecting an inner box in a drawing shape to be filled, wherein the inner box includes an area that is to be left unfilled;determining whether the detected inner box is big enough for pre-clipping; andconfirming that an outer clip path contains the detected inner box.2. The method of claim 1 , wherein an inner clip path is a rectangle claim 1 , and wherein detecting the inner box comprises detecting that the inner box is the rectangle.3. The method of claim 1 , wherein an inner clip path consists of polygon lines claim 1 , and wherein detecting the inner box comprises:calculating the center point for the polygon lines;comparing each polygon point with the calculated center point; anddetermining at least two corner coordinates of the inner box.4. The method of claim 3 , wherein calculating the center point for the polygon lines comprises:{'b': 0', '1, 'calculating an X-coordinate of the center point as Xcenter=(X+X+ . . . +Xn−1)/n, where n is a point number of polygon lines; and'}{'b': 0', '1, 'calculating a Y-coordinate of the center point as Ycenter=(Y+Y+ . . . +Yn−1)/n.'}5. The method of claim 3 , wherein comparing each polygon point with the calculated center point for the ...

VOXELS SPARSE REPRESENTATION

Embodiments described herein provide an apparatus comprising a processor to project voxels from a point cloud data set into an n-DoF space, and define successively less granular supervoxels at successively higher layer of abstraction in a view of the point cloud data set, and a memory communicatively coupled to the processor. Other embodiments may be described and claimed. 1. A method , comprising:projecting voxels from a point cloud data set into an n-DoF space; anddefining successively less granular supervoxels at successively higher layer of abstraction in a view of the point cloud data set.2. The method of claim 1 , further comprising:combining voxels from the point cloud data set into a first set of supervoxels at a first layer of abstraction; andcombining supervoxels from the first set of supervoxels to form a second set of supervoxels at a second layer of abstraction.3. The method of claim 1 , further comprising:defining a first abstraction layer;implementing a labeling process to associate voxels in the point cloud data set into one or more clusters based upon characteristics of the voxels; andassigning labels to the one or more clusters.4. The method of claim 3 , further comprising:selecting a cluster from the plurality of clusters;marking voxels within predefined Euclidian distance of a center of the cluster;calculating a set of expected values of color data, normal data, and coordinate data for the voxels; anddefining a set of supervoxels according to the set of expected values of color data, normal data, and coordinate data for the voxels.5. The method of claim 4 , further comprising:calculating a set of covariance values of the color data, normal data, and coordinate data for the voxels; anddefining dimensions for the set of supervoxels according to the covariance values of the color data, normal data, and coordinate data for the voxels.6. The method of claim 5 , further comprising:removing the labels from the one or more clusters; andprojecting the set ...

OBJECT IDENTIFICATION IN DIGITAL IMAGES

In an example, a digital image comprising a representation of multiple physical objects is received at a client computer. The digital image is copied into a temporary canvas. The digital image is then analyzed to identify a plurality of potential object areas, each of the potential object areas having pixels with colors similar to the other pixels within the potential object area. A minimum bounding region for each of the identified potential object areas is identified, the minimum bounding region being a smallest region of a particular shape that bounds the corresponding potential object area. The pixels within a selected minimum bounding region are cropped from the digital image. The pixels within the selected minimum bounding region are then sent to an object recognition service on a server to identify an object represented by the pixels within the selected minimum bounding region. 120-. (canceled)21. A system comprising:at least one processor; and receiving, at a client computer, a digital image comprising a representation of multiple physical objects;', 'identifying, based on pixels extracted at the client computer, a separate minimum bounding region for each of a plurality of potential object areas, the separate minimum bounding region being a smallest region of a particular shape that bounds the corresponding potential object area;', 'cropping, at the client computer, pixels within a selected minimum bounding region from the digital image, wherein the cropping includes removing the pixels from the digital image;', 'sending, at the client computer and to a publication system including an object recognition service, the cropped pixels within the selected minimum bounding region to identify an object represented by the cropped pixels within the selected minimum bounding region; and', 'receiving, at the client computer and in response to sending the cropped pixels within the selected minimum bounding region, a listing from the publication system, the listing ...

GENERATING REFLECTIONS WITHIN IMAGES

Techniques for propagating a reflection of an object. In an example, a method includes receiving an input image comprising a first reflection of a first object on a reflective surface. The method further includes generating a second reflection for a second object in the input image. The second reflection is a reflection of the second object on the reflective surface. The method includes adding the second reflection to the input image. The method includes outputting a modified image comprising the first object, first reflection, the second object, and the second reflection. 1. A computer-implemented method comprising:accessing, by a processor, an input image comprising a first object, a first reflection of the first object on a reflective surface in the input image, and a second object;analyzing, by the processor, the first object and the first reflection to compute a reflection parameter, wherein the analyzing comprises computing the reflection parameter based upon a shape of the first object and a shape of the first reflection;generating, by the processor and using the reflection parameter, a second reflection of the second object on the reflective surface, wherein the generating comprises applying the reflection parameter to the second object;adding, by the processor, the second reflection to the input image; andoutputting, by the processor, a modified image comprising the first object, the first reflection, the second object, and the second reflection.2. The method of claim 1 , further comprising:applying a noise-reduction filter to the input image; andidentifying, from the input image, the first object and the first reflection by performing a segmentation on the input image.3. The method of claim 1 , further comprising:receiving a signal to generate the second reflection; anddetermining that the first object and the first reflection are to be used as references for generating the second reflection for the second object.4. The method of claim 1 , wherein the ...

PRE-SEGMENT POINT CLOUD DATA TO RUN REAL-TIME SHAPE EXTRACTION FASTER

A method, apparatus, system, and computer readable storage medium provide the ability to pre-segment point cloud data. Point cloud data is obtained and segmented. Based on the segment information, a determination is made regarding points needed for shape extraction. Needed points are fetched and used to extract shapes. The extracted shapes are used to cull points from the point cloud data. 1. A computer-implemented method for pre-segmenting point cloud data , the method comprising the steps of:obtaining point cloud data;segmenting the point cloud data resulting in segment information;determining, based on the segment information, points needed for shape extraction;fetching the points needed for shape extraction;extracting one or more shapes based on the fetched points; andutilizing the extracted one or more shapes to cull points from the point cloud data.2. The computer-implemented method of claim 1 , further comprising:creating and fitting planes on the point cloud data;culling the point cloud data based on the planes; andrendering the culled point cloud data; andwherein the pre-segmenting steps are performed during an indexing of the point cloud data.3. The computer-implemented method of claim 1 , wherein the segmenting comprises:calculating a surface normal of each point in the point cloud data using a spatial neighborhood of each point;refining the surface normals by assigning each point to a surface type and updating normal information using neighbors that belong to a same surface type; andclustering the points in the point cloud data into segments based on spatial proximity and surface normal angles.4. The computer-implemented method of claim 1 , wherein the segment information comprises:the segment identification; anda bounding box.5. The computer-implemented method of claim 4 , wherein the segment information further comprises homogeneity information and planarity.6. The computer-implemented method of claim 1 , wherein the extracting comprises a least ...

Methods and systems of maintaining lost object trackers in video analytics

Techniques and systems are provided for maintaining lost blob trackers for one or more video frames. In some examples, one or more blob trackers maintained for a sequence of video frames are identified. The one or more blob trackers are associated with one or more blobs of the sequence of video frames. A transition of a blob tracker from a first type of tracker to a lost tracker is detected at a first video frame. For example, the blob tracker can be transitioned from the first type of tracker to the lost tracker when a blob for which the blob tracker was associated with in a previous frame is not detected in the first video frame. A recovery duration is determined for the lost tracker at the first video frame. For one or more subsequent video frames obtained after the first video frame, the lost tracker is removed from the one or more blob trackers maintained for the sequence of video frames when a lost duration for the lost tracker is greater than the recovery duration. The blob tracker can be transitioned back to the first type of tracker if the lost tracker is associated with a blob in a subsequent video frame prior to expiration of the recovery duration. Trackers and associated blobs are output as identified blob tracker-blob pairs when the trackers are converted from new trackers to trackers of the first type.

ADAPTIVE BOUNDING BOX MERGE METHOD IN BLOB ANALYSIS FOR VIDEO ANALYTICS

Provided are methods, apparatuses, and computer-readable medium for content-adaptive bounding box merging. A system using content-adaptive bounding box merging can adapt its merging criteria according to the objects typically present in a scene. When two bounding boxes overlap, the content-adaptive merge engine can consider the overlap ratio, and compare the merged bounding box against a minimum object size. The minimum object size can be adapted to the size of the blobs detected in the scene. When two bounding boxes do not overlap, the system can consider the horizontal and vertical distances between the bounding boxes. The system can further compare the distances against content-adaptive thresholds. Using a content-adaptive bounding box merge engine, a video content analysis system may be able to more accurately merge (or not merge) bounding boxes and their associated blobs. 1. A method for merging bounding boxes , comprising:determining a candidate merged bounding box for a first bounding box and a second bounding box, wherein the first bounding box is associated with a first blob, wherein the first blob includes pixels of at least a portion of a first foreground object in a video frame, wherein the second bounding box is associated with a second blob, wherein the second blob includes pixels of at least a portion of a second foreground object in the video frame, and wherein the candidate merged bounding box includes the first blob and the second blob;determining a size of the candidate merged bounding box;comparing the size of the candidate merged bounding box against a size threshold; anddetermining to merge the first bounding box and the second bounding box based on the size of the candidate merged bounding box being less than the size threshold.2. The method of claim 1 , further comprising:determining that the first bounding box and the second bounding box have an intersecting region and a non-intersecting region;determining a ratio between an area of the non- ...

Method and system for identifying, tracking, and collecting data on a person of interest

A system may be configured to reliably identify, track, and collect data on a person of interest. In some aspects, the system may detect a facial representation within an image, generate a bounding box corresponding to the facial representation, determine an enhanced facial representation based at least in part on the bounding box and an image enhancement pipeline, and extract a plurality of facial features from the enhanced facial representation. Further, the system may determine quality information based on the plurality of facial features, determine that the quality information is greater than a threshold, and store the plurality of facial features in a data structure. In some aspects, the facial features may be used to identify red shoppers and green shoppers in a controlled environment.

VEHICLE PARKING MONITORING SYSTEMS AND METHODS

Vehicle parking monitoring systems and methods are disclosed herein. An example method can include receiving images from a camera of a parking spot, each of the images being time stamped, determining presence of a vehicle in the images, placing a bounding area around a region of interest of the vehicle, the region of interest including no personally identifiable information, retaining the bounding area and discarding a remainder of the images, and determining when the vehicle is no longer present based on a change in the bounding area of the images. 1. A method , comprising:receiving images of a parking spot from a camera, wherein each of the images is time stamped;determining presence of a vehicle in the images;placing a bounding area around a region of interest of the vehicle, wherein the region of interest includes no personally identifiable information;retaining the bounding area and discarding a remainder of the images; anddetermining, based on a change in the bounding area of the images, that the vehicle is no longer present.2. The method according to claim 1 , further comprising detecting motion in the bounding area between temporally adjacent images.3. The method according to claim 1 , further comprising:determining that the vehicle has not moved within an allotted time period; andtransmitting, based on the vehicle not moving within the allotted time period, an alert to a service provider.4. The method according to claim 1 , further comprising detecting claim 1 , based on output of an accelerometer associated with the camera claim 1 , that the camera has been tampered.5. The method according to claim 1 , further comprising applying error correction to compensate for movement of the camera.6. The method according to claim 1 , further comprising applying error correction to compensate for ambient weather conditions.7. A system claim 1 , comprising:an imaging device that obtains images of a parking spot, wherein each of the images is time stamped; and determine ...

SYSTEM AND METHOD FOR TRACKING OBJECTS USING USING EXPANDED BOUNDING BOX FACTORS

Systems and methods of tracking objects around an ego vehicle, using state estimation of the objects, where state estimation may include an estimate of a position, velocity, acceleration and bounding box size of the tracked object are disclosed. An object may be detected at one or more poses over time. For each pose, a first bounding box may be generated. The largest bounding box may be determined. For each pose, the first bounding box may be expanded to an expanded bounding box. The expanded bounding box may match the largest bounding box. For each pose, the center of the expanded bounding box may be determined. The object may be tracked according to the center of the expanded bounding box.

INFERRING INTENT USING COMPUTER VISION

A system trains a model to infer an intent of an entity. The model includes one or more sensors to obtain frames of data, one or more processors, and a memory storing instructions that, when executed by the one or more processors, cause the system to perform steps. A first step includes determining, in each frame of the frames, one or more bounding regions, each of the bounding regions enclosing an entity. A second step includes identifying a common entity, the common entity being present in bounding regions corresponding to a plurality of the frames. A third step includes associating the common entity across the frames. A fourth step includes training a model to infer an intent of the common entity based on data outside of the bounding regions. 1. A system configured to train a model to infer an intent of an entity , comprising:one or more sensors configured to obtain frames of data;one or more processors; and: determining, in each frame of the frames, one or more bounding regions, each of the bounding regions enclosing an entity;', 'across the frames, identifying a common entity, the common entity being present in bounding regions corresponding to a plurality of the frames;', 'associating the common entity across the frames;', 'training a model to infer an intent associated with the common entity based on data outside of the bounding regions., 'a memory storing instructions that, when executed by the one or more processors, cause the system to perform2. The system of claim 1 , wherein the training the model is based on an output generated from an other model.3. The system of claim 2 , wherein the output generated from the other model comprises a segmentation output.4. The system of claim 3 , wherein the segmentation output comprises a semantic segmentation output or an instance segmentation output.5. The system of claim 4 , wherein the segmentation output comprises a semantic segmentation output that trains the model to predict a category or classification ...

SYSTEM FOR MANAGING CONSUMER PACKAGED GOODS

A system for managing consumer packaged goods (CPGs) is disclosed. The system includes a front end and a back end. The front end includes a mobile, airborne platform equipped with a digital image capturing device, and a wireless communications device. The backend is in communication with the front end via the wireless communications device, and includes a combinational convolutional neural network which derives models based on input data, a convolutional neural network which generates perception scoring utilizing input put from said combinational convolutional neural network, and a recurrent neural network which makes behavior predictions based on input from said convolutional neural network. The front end captures images of CPGs on a shelf, wherein the CPGs are subject to depletion over time, and wherein the backend generates predictions regarding the depletion state of the CPGs on the shelf as a function of time.

SYSTEMS AND METHODS FOR TEXTURE MAPPING FOR A 3D MODEL

Systems and methods for constructing a three-dimensional virtual reality model of a property are disclosed. An exemplary system includes a storage device configured to store a structure mesh of the three-dimensional virtual reality model and texture images captured of the property. The system further includes at least one processor configured to extract a planar area from the structure mesh of the three-dimensional virtual reality model. The at least one processor is further configured to obtain initial texture patches for the planar area from the texture images, determine a target texture patch for the planar area from the initial texture patches projected to an imaging plane of a virtual camera, and add the target texture patch to a texture atlas. The at least one processor is also configured to construct the three-dimensional virtual reality model by mapping the texture atlas to the structure mesh of the three-dimensional virtual reality model.

Image Processing Techniques

Techniques are described that can delay or even prevent use of memory to store triangles associated with tiles as well as processing resources associated with vertex shading and binning triangles. The techniques can also provide better load balancing among a set of cores, and hence provide better performance. A bounding volume is generated to represent a geometry group. Culling takes place to determine whether a geometry group is to have triangles rendered. Vertex shading and association of triangles with tiles can be performed across multiple cores in parallel. Processing resources are allocated for rasterizing tiles whose triangles have been vertex shaded and binned over tiles whose triangles have yet to be vertex shaded and binned. Rasterization of triangles of different tiles can be performed by multiple cores in parallel. 1. A system comprising:a processor having multiple cores;a wireless interface communicatively coupled to the processor;a display device communicatively coupled to the processor; and determine a first bounding volume for a first geometry group;', 'determine a second bounding volume for a second geometry group during a time span overlapping with a time span to determine the first bounding volume;', 'assign the first geometry group to one or more geometry group bins based on the first bounding volume;', 'perform back-end processing on a tile identified as ready for back-end processing; and', 'perform mid-end processing in response to not identifying a tile ready for back-end processing during a time span overlapping with the time span to perform back-end processing., 'a memory communicatively coupled to the processor, wherein at least one of the cores is configured to2. The system of claim 1 , wherein to determine the first bounding volume and to determine the second bounding volume claim 1 , at least one of the cores is to:apply bounded arithmetic techniques to determine a bounding volume for each geometry group.3. The system of claim 1 , ...