Настройки

Укажите год
-

Небесная энциклопедия

Космические корабли и станции, автоматические КА и методы их проектирования, бортовые комплексы управления, системы и средства жизнеобеспечения, особенности технологии производства ракетно-космических систем

Подробнее
-

Мониторинг СМИ

Мониторинг СМИ и социальных сетей. Сканирование интернета, новостных сайтов, специализированных контентных площадок на базе мессенджеров. Гибкие настройки фильтров и первоначальных источников.

Подробнее

Форма поиска

Поддерживает ввод нескольких поисковых фраз (по одной на строку). При поиске обеспечивает поддержку морфологии русского и английского языка
Ведите корректный номера.
Ведите корректный номера.
Ведите корректный номера.
Ведите корректный номера.
Укажите год
Укажите год
Применить Всего найдено 23. Отображено 22.
14-09-2017 дата публикации

SYSTEMS AND METHODS FOR NORMALIZING AN IMAGE

Номер: US20170262962A1
Автор: Mahdi Rad, Markus Oberweger, Vincent Lepetit, RAD MAHDI, OBERWEGER MARKUS, LEPETIT VINCENT, Rad Mahdi, Oberweger Markus, Lepetit Vincent
Принадлежит:

A method for normalizing an image by an electronic device is described. The method includes obtaining an image including a target object. The method also includes determining a set of windows of the image. The method further includes, for each window of the set of windows of the image, predicting parameters of an illumination normalization model adapted to the window using a first convolutional neural network (CNN), and applying the illumination normalization model to the window to produce a normalized window. 1. A method for normalizing an image by an electronic device , the method comprising:obtaining an image comprising a target object;determining a set of windows of the image; predicting parameters of an illumination normalization model adapted to the window using a first convolutional neural network (CNN); and', 'applying the illumination normalization model to the window to produce a normalized window., 'for each window of the set of windows of the image2. The method of claim 1 , further comprising claim 1 , for each window of the set of windows of the image claim 1 , analyzing the normalized window with a second CNN for scoring the normalized window.3. The method of claim 2 , wherein the first CNN and the second CNN are jointly trained CNNs.4. The method of claim 2 , further comprising detecting the target object in a normalized window with a largest score.5. The method of claim 2 , wherein the first CNN is a normalizer CNN and the second CNN is a detector CNN.6. The method of claim 2 , wherein the first CNN is trained for predicting parameters of the illumination normalization model based on a set of training images claim 2 , wherein a subset of the set of training images comprises a training target object claim 2 , and wherein the second CNN is trained for detecting the training target object based on the set of training images and the illumination normalization model.7. The method of claim 2 , wherein the first CNN and the second CNN are trained based on ...

Подробнее
Номер записи: 1
14-03-2019 дата публикации

POSE DETERMINATION WITH SEMANTIC SEGMENTATION

Номер: US20190080467A1
Автор: Arth Clemens, Hirzer Martin, Lepetit Vincent, ROTH Peter Michael
Принадлежит:

A method determines a pose of an image capture device. The method includes accessing an image of a scene captured by the image capture device. A semantic segmentation of the image is performed, to generate a segmented image. An initial pose of the image capture device is generated using a three-dimensional (3D) tracker. A plurality of 3D renderings of the scene are generated, each of the plurality of 3D renderings corresponding to one of a plurality of poses chosen based on the initial pose. A pose is selected from the plurality of poses, such that the 3D rendering corresponding to the selected pose aligns with the segmented image. 1. A method of determining a pose of an image capture device , comprising:accessing an image of a scene captured by the image capture device;performing a semantic segmentation of the image, to generate a segmented image;generating an initial pose of the image capture device using a three-dimensional (3D) tracker;generating a plurality of 3D renderings of the scene, each of the plurality of 3D renderings corresponding to one of a plurality of poses chosen based on the initial pose; andselecting a pose from the plurality of poses, such that the 3D rendering corresponding to the selected pose aligns with the segmented image.2. The method of determining a pose according to claim 1 , further comprising updating the 3D tracker based on the selected pose.3. The method of determining a pose according to claim 1 , further comprising:capturing the image using the image capture device; andrectifying the image before the semantic segmentation.4. The method of determining a pose according to claim 1 , wherein generating the initial pose includes using calibration data claim 1 , position data or motion data from one or more sensors.5. The method of claim 4 , wherein the plurality of poses includes the initial pose claim 4 , and the calibration data claim 4 , position data or motion data are used to determine a pose search space containing the plurality ...

Подробнее
Номер записи: 2
05-04-2018 дата публикации

Method, System, and Device for Learned Invariant Feature Transform for Computer Images

Номер: US20180096224A1
Автор: Fua Pascal, Lepetit Vincent, Trulls Eduard Fortuny, YI Kwang Moo
Принадлежит:

A method for training a feature detector of an image processing device, including the steps of detecting features in the image to generate a score map, computing a center of mass on the score map to generate a location, extracting a patch from the image at the location by a first spatial transformer, estimating an orientation of the patch, rotating the patch in accordance with the patch orientation with a second spatial transformer, and describing the rotated patch to create a description vector. 1. A method for training a feature detector of an image processor , the method comprising the steps of:detecting features in the image to generate a score map;computing a center of mass on the score map to generate a location;extracting a patch from the image at the location by a first spatial transformer;estimating an orientation of the patch;rotating the patch in accordance with the patch orientation with a second spatial transformer; anddescribing the rotated patch to create a description vector.2. The method according to claim 1 , wherein in the step of computing claim 1 , a softargmax is performed on the score map to generate the location of a single potential feature point of the image.3. The method according to claim 1 , wherein in the step of estimating claim 1 , an orientation estimator is used that estimates the orientation that minimize distances between description vectors for different views of same 3D points.4. The method according to claim 1 , wherein in the step of detecting claim 1 , a convolution layer is used to generate the score map followed by a piecewise linear activation function.5. The method according to claim 1 , wherein in the step of describing claim 1 , a descriptor is used that includes three convolutional layers followed by hyperbolic tangent units claim 1 , 12 pooling claim 1 , and local subtractive normalization.6. The method according to claim 5 , wherein the descriptor is trained by minimizing a sum of a loss for pairs of corresponding ...

Подробнее
Номер записи: 3
17-05-2018 дата публикации

METHODS AND SYSTEMS OF PERFORMING OBJECT POSE ESTIMATION

Номер: US20180137644A1
Автор: Lepetit Vincent, Oberweger Markus, Rad Mahdi
Принадлежит:

Techniques are provided for estimating a three-dimensional pose of an object. An image including the object can be obtained, and a plurality of two-dimensional (2D) projections of a three-dimensional bounding (3D) box of the object in the image can be determined. The plurality of 2D projections of the 3D bounding box can be determined by applying a trained regressor to the image. The trained regressor is trained to predict two-dimensional projections of the 3D bounding box of the object in a plurality of poses, based on a plurality of training images. The three-dimensional pose of the object is estimated using the plurality of 2D projections of the 3D bounding box. 1. A method of estimating a three-dimensional pose of an object , the method comprising:obtaining an image including the object;determining a plurality of two-dimensional projections of a three-dimensional bounding box of the object, the plurality of two-dimensional projections of the three-dimensional bounding box being determined by applying a trained regressor to the image, wherein the trained regressor is trained to predict two-dimensional projections of the three-dimensional bounding box of the object in a plurality of poses; andestimating the three-dimensional pose of the object using the plurality of two-dimensional projections of the three-dimensional bounding box.2. The method of claim 1 , wherein the regressor is a convolutional neural network.3. The method of claim 1 , further comprising determining an image patch in the image claim 1 , wherein the image patch is centered on the object claim 1 , and wherein the plurality of two-dimensional projections of the three-dimensional bounding box are determined by applying the trained regressor to the image patch.4. The method of claim 3 , wherein determining the image patch includes:segmenting the image into an object area map of the image, the object area map including a grid of object area blocks;determining a respective probability for each of the ...

Подробнее
Номер записи: 4
16-05-2019 дата публикации

POSE ESTIMATION AND MODEL RETRIEVAL FOR OBJECTS IN IMAGES

Номер: US20190147221A1
Автор: GRABNER Alexander, Lepetit Vincent, ROTH Peter Michael
Принадлежит:

Techniques are provided for selecting a three-dimensional model. An input image including an object can be obtained, and a pose of the object in the input image can be determined. One or more candidate three-dimensional models representing one or more objects in the determined pose can be obtained. From the one or more candidate three-dimensional models, a candidate three-dimensional model can be determined to represent the object in the input image. 1. A method of selecting a three-dimensional model , the method comprising:obtaining an input image including an object;determining a pose of the object in the input image;obtaining one or more candidate three-dimensional models representing one or more objects in the determined pose; anddetermining, from the one or more candidate three-dimensional models, a candidate three-dimensional model to represent the object in the input image.2. The method of claim 1 , further comprising generating an output image based on the candidate three-dimensional model and the input image.3. The method of claim 1 , further comprising:receiving a user input to manipulate the candidate three-dimensional model; andadjusting one or more of a pose or a location of the candidate three-dimensional model in an output image based on the user input.4. The method of claim 1 , further comprising:obtaining an additional input image, the additional input image including the object in one or more of a different pose or a different location than a pose or location of the object in the input image; andadjusting one or more of a pose or a location of the candidate three-dimensional model in an output image based on a difference between the pose or location of the object in the additional input image and the pose or location of the object in the input image.5. The method of claim 1 , wherein obtaining the one or more three-dimensional models representing the one or more objects includes:obtaining a plurality of three-dimensional models representing a ...

Подробнее
Номер записи: 5
22-09-2022 дата публикации

KEYPOINT-BASED SAMPLING FOR POSE ESTIMATION

Номер: US20220301304A1
Автор: Arth Clemens, HAMPALI Shreyas, Lepetit Vincent
Принадлежит:

Systems and techniques are provided for determining one or more poses of one or more objects. For example, a process can include determining, using a machine learning system, a plurality of keypoints from an image. The plurality of keypoints are associated with at least one object in the image. The process can include determining a plurality of features from the machine learning system based on the plurality of keypoints. The process can include classifying the plurality of features into a plurality of joint types. The process can include determining pose parameters for the at least one object based on the plurality of joint types. 1. An apparatus for determining one or more poses of one or more objects , comprising:at least one memory; and determine, using a machine learning system, a plurality of keypoints from an image, the plurality of keypoints being associated with at least one object in the image;', 'determine a plurality of features from the machine learning system based on the plurality of keypoints;', 'classify the plurality of features into a plurality of joint types; and', 'determine pose parameters for the at least one object based on the plurality of joint types., 'at least one processor coupled to the at least one memory, the at least one processor configured to2. The apparatus of claim 1 , wherein the at least one object includes at least one hand.3. The apparatus of claim 1 , wherein the at least one object includes two objects claim 1 , wherein the plurality of keypoints includes keypoints for the two objects claim 1 , and wherein the pose parameters include pose parameters for the two objects.4. The apparatus of claim 1 , wherein the at least one object includes two hands claim 1 , wherein the plurality of keypoints includes keypoints for the two hands claim 1 , and wherein the pose parameters include pose parameters for the two hands.5. The apparatus of claim 1 , wherein the at least one object includes a single hand claim 1 , and wherein the at ...

Подробнее
Номер записи: 6
11-06-2020 дата публикации

Systems and methods for three-dimensional pose determination

Номер: US20200184668A1
Автор: Mahdi RAD, Markus OBERWEGER, Vincent Lepetit
Принадлежит: Qualcomm Inc, Qualcomm Technologies Inc

A method is described. The method includes mapping features extracted from an unannotated red-green-blue (RGB) image of the object to a depth domain. The method further includes determining a three-dimensional (3D) pose of the object by providing the features mapped from the unannotated RGB image of the object to the depth domain to a trained pose estimator network.

Подробнее
Номер записи: 7
20-09-2018 дата публикации

THREE-DIMENSIONAL POSE ESTIMATION OF SYMMETRICAL OBJECTS

Номер: US20180268601A1
Автор: Lepetit Vincent, Oberweger Markus, Rad Mahdi
Принадлежит:

The present disclosure describes methods, apparatuses, and non-transitory computer-readable mediums for estimating a three-dimensional (“3D”) pose of an object from a two-dimensional (“2D”) input image which contains the object. Particularly, certain aspects of the disclosure are concerned with 3D pose estimation of a symmetric or nearly-symmetric object. An image or a patch of an image includes the object. A classifier is used to determine whether a rotation angle of the object in the image or the patch of the image is within a first predetermined range. In response to a determination that the rotation angle is within the first predetermined range, a mirror image of the object is determined. Two-dimensional (2D) projections of a three-dimensional (3D) bounding box of the object are determined by applying a trained regressor to the mirror image of the object in the image or the patch of the image. The 3D pose of the object is estimated based on the 2D projections. 1. A method for estimating a three-dimensional (“3D”) pose of an object having an angle of rotational symmetry , the method comprising:obtaining an image or a patch of an image which includes the object;determining, via a classifier, whether a rotation angle of the object in the image or the patch of the image is within a first predetermined range; and determining a mirror image of the object;', 'determining two-dimensional (2D) projections of a three-dimensional (3D) bounding box of the object by applying a trained regressor to the mirror image of the object in the image or the patch of the image; and', 'estimating the 3D pose of the object based on the 2D projections., 'in response to a determination that the rotation angle is within the first predetermined range2. The method of wherein the trained regressor is trained using a plurality of training images claim 1 , wherein each of the training images is taken from a different point of view and includes the object with a different pose from a plurality of ...

Подробнее
Номер записи: 8
02-11-2017 дата публикации

Method, System and Device for Direct Prediction of 3D Body Poses from Motion Compensated Sequence

Номер: US20170316578A1
Автор: Fua Pascal, Lepetit Vincent, Rozantsev Artem, Tekin Bugra
Принадлежит:

A method for predicting three-dimensional body poses from image sequences of an object, the method performed on a processor of a computer having memory, the method including the steps of accessing the image sequences from the memory, finding bounding boxes around the object in consecutive frames of the image sequence, compensating motion of the object to form spatio-temporal volumes, and learning a mapping from the spatio-temporal volumes to a three-dimensional body pose in a central frame based on a mapping function. 1. A method for predicting three-dimensional body poses from image sequences of an object , the method performed on a processor of a computer having memory , the method comprising the steps of:accessing the image sequences from the memory;finding bounding boxes around the object in consecutive frames of the image sequence;compensating motion of the object to form spatio-temporal volumes; andlearning a mapping from the spatio-temporal volumes to a three-dimensional body pose in a central frame based on a mapping function.2. The method according to claim 1 , wherein the step of compensating motion includes centering the object in consecutive frames.3. The method according to claim 1 , wherein the mapping function uses a feature vector from the spatio-temporal volumes based on a histogram of oriented gradients (HOG) descriptor.4. The method according to claim 3 , wherein the HOG descriptor uses volume cells having different cell sizes.5. The method according to claim 4 , wherein in the step of compensating motion claim 4 , convolutional neural net regressors are trained to estimate a shift of the object from a center of the bounding boxes.6. The method according to claim 1 , wherein the object is a living being.7. A device for predicting three-dimensional body poses from image sequences of an object claim 1 , the device including a processor having access to a memory claim 1 , the processor configured to:access the image sequences from the memory;find ...

Подробнее
Номер записи: 9
23-11-2017 дата публикации

PREDICTOR-CORRECTOR BASED POSE DETECTION

Номер: US20170337690A1
Автор: Arth Clemens, Lepetit Vincent, Wohlhart Paul
Принадлежит:

A computer-implemented method, apparatus, computer readable medium and mobile device for determining a 6DOF pose from an input image. The process of determining 6DOF pose may include processing an input image to create one or more static representations of the input image, creating a dynamic representation of the input image from an estimated 6DOF pose and a 2.5D reference map, and measuring correlation between the dynamic representation and the one or more static representations of the input image. The estimated 6DOF may be iteratively adjusted according to the measured correlation error until a final adjusted dynamic representation meets an output threshold. 1. A computer-implemented method for determining a 6DOF camera pose , the method comprising:creating one or more static representations of an input image;creating a dynamic representation of the input image from an estimated 6DOF pose and a 2.5D reference map;measuring a correlation error between the dynamic representation and the one or more static representations;adjusting the estimated 6DOF pose according to the correlation error;updating the dynamic representation according to the adjusted 6DOF pose; andoutputting the adjusted 6DOF pose in response to meeting an output threshold.2. The computer-implemented method of claim 1 , wherein the one or more static representations includes one or more of:class segments determined by segmenting parts of the input image into respective classes,planar structures determined by segmenting the input image into planar structures,line features determined by segmenting the input image into lines,a depth map determined by estimating depth for each pixel within the image, orany combination thereof.3. The computer-implemented method of claim 1 , wherein the output threshold is met when:the correlation error is within an error threshold,an iteration count is met, ora combination thereof.4. The computer-implemented method of claim 1 , further comprising:iteratively updating the ...

Подробнее
Номер записи: 10
24-12-2015 дата публикации

ZERO-BASELINE 3D MAP INITIALIZATION

Номер: US20150371440A1
Автор: Arth Clemens, Lepetit Vincent, Pirchheim Christian, Schmalstieg Dieter, Ventura Jonathan
Принадлежит:

A computer-implemented method, apparatus, computer readable medium and mobile device for initializing a 3-Dimensional (3D) map may include obtaining, from a camera, a single image of an urban outdoor scene and estimating an initial pose of the camera. An untextured model of a geographic region may be obtained. Line features from the single image may be extracted and the orientation may be determined with respect to the untextured model and using the extracted line features, the orientation of the camera in 3 Degrees of Freedom (3DOF). In response to determining the orientation of the camera, a translation in 3DOF with respect to the untextured model may be determined using the extracted line features. The 3D map may be initialized based on the determined orientation and translation. 1. A computer-implemented method of initializing a 3-Dimensional (3D) map , the method comprising:obtaining, from a camera, a single image of an urban outdoor scene;estimating, from one or more device sensors, an initial pose of the camera;obtaining, based at least in part on the estimated initial pose, an untextured model of a geographic region that includes the urban outdoor scene;extracting a plurality of line features from the single image;determining, with respect to the untextured model and using the extracted line features, the orientation of the camera in 3 Degrees of Freedom (3DOF);determining, in response to determining the orientation of the camera, a translation in 3DOF with respect to the untextured model and using the extracted line features; andinitializing the 3D map based on the determined orientation and translation.2. The computer-implemented method of claim 1 , wherein the one or more device sensors include: a satellite positioning system claim 1 , accelerometer claim 1 , magnetometer claim 1 , gyroscope claim 1 , or any combination thereof.3. The computer-implemented method of claim 1 , wherein the untextured model includes a 2.5D topographical map with building ...

Подробнее
Номер записи: 11
17-05-2018 дата публикации

Methods and systems of performing object pose estimation

Номер: WO2018089163A1
Автор: Mahdi RAD, Markus OBERWEGER, Vincent Lepetit
Принадлежит: QUALCOMM INCORPORATED

Techniques are provided for estimating a three-dimensional pose of an object. An image including the object can be obtained, and a plurality of two-dimensional (2D) projections of a three-dimensional bounding (3D) box of the object in the image can be determined. The plurality of 2D projections of the 3D bounding box can be determined by applying a trained regressor to the image. The trained regressor is trained to predict twodimensional projections of the 3D bounding box of the object in a plurality of poses, based on a plurality of training images. The three-dimensional pose of the object is estimated using the plurality of 2D projections of the 3D bounding box.

Подробнее
Номер записи: 12
20-05-2021 дата публикации

Layout estimation using planes

Номер: US20210150805A1
Автор: Friedrich Fraundorfer, Sinisa STEKOVIC, Vincent Lepetit
Принадлежит: Qualcomm Inc, Technische Universitaet Graz

Techniques are provided for determining one or more environmental layouts. For example, one or more planes can be detected in an input image of an environment. The one or more planes correspond to one or more objects in the input image. One or more three-dimensional parameters of the one or more planes can be determined. One or more polygons can be determined using the one or more planes and the one or more three-dimensional parameters of the one or more planes. A three-dimensional layout of the environment can be determined based on the one or more polygons.

Подробнее
Номер записи: 13
03-10-2023 дата публикации

Amostragem com base em pontos-chaves para estimação de pose

Номер: BR112023018094A2
Автор: Clemens Arth, Shreyas HAMPALI, Vincent Lepetit
Принадлежит: Qualcomm Technologies Inc

amostragem com base em pontos-chaves para estimação de pose. sistemas e técnicas são fornecidos para determinar uma ou mais poses de um ou mais objetos. por exemplo, um processo pode incluir determinar, com utilização de um sistema de aprendizado de máquina, uma pluralidade de pontos-chaves proveniente de uma imagem. a pluralidade de pontos-chaves está associada a pelo menos um objeto na imagem. o processo pode incluir determinar uma pluralidade de características do sistema de aprendizado de máquina com base na pluralidade de pontos-chaves. o processo pode incluir classificar a pluralidade de características em uma pluralidade de tipos de articulação. o processo pode incluir determinar parâmetros de pose para pelo menos um objeto com base na pluralidade de tipos de articulação.

Подробнее
Номер записи: 14
14-02-2024 дата публикации

Systems and methods for grasping objects like humans using robot grippers

Номер: EP4321309A1
Автор: Philippe Weinzaepfel, Romain Bregier, Vincent Lepetit, Yumin DU
Принадлежит: Ecole Nationale des Ponts et Chaussess ENPC, Naver Corp, Naver Labs Corp

A computer-implemented method, comprising:based on a demonstration of a human hand grasping an object, determining (a) a first vector that is normal to a palm of the human hand, (b) a second vector that is parallel to the palm of the human hand, and (c) a position of the human hand;determining (a) a third vector that is normal to a palm of a gripper of a robot, (b) a fourth vector that is parallel to the palm of the gripper of the robot, and (c) a present position of the gripper;(1) actuating one or more actuators of the robot and moving the gripper when the gripper is open such that the present position of the gripper is at the position of the human hand, the third vector is aligned with the first vector, and the fourth vector is aligned with the second vector;after (1), (2) actuating one or more actuators of the robot and closing fingers of the gripper based on minimizing an object penetration loss; andwhen the object penetration loss is minimized, (3) actuating one or more actuators of the robot and actuating the fingers of the gripper to minimize a total loss determined based on the object penetration loss and at least one other loss.

Подробнее
Номер записи: 15
20-09-2023 дата публикации

Scene layout estimation

Номер: EP4244824A1
Автор: Friedrich Fraundorfer, Shreyas HAMPALI, Sinisa STEKOVIC, Vincent Lepetit
Принадлежит: Qualcomm Technologies Inc

Systems and techniques are provided for determining environmental layouts. For example, based on one or more images of an environment and depth information associated with the one or more images, a set of candidate layouts and a set of candidate objects corresponding to the environment can be detected. The set of candidate layouts and set of candidate objects can be organized as a structured tree. For instance, a structured tree can be generated including nodes corresponding to the set of candidate layouts and the set of candidate objects. A combination of objects and layouts can be selected in the structured tree (e.g., based on a search of the structured tree, such as using a Monte-Carlo Tree Search (MCTS) algorithm or adapted MCTS algorithm). A three-dimensional (3D) layout of the environment can be determined based on the combination of objects and layouts in the structured tree.

Подробнее
Номер записи: 16
24-10-2023 дата публикации

Scene layout estimation

Номер: US11797724B2
Автор: Friedrich Fraundorfer, Shreyas HAMPALI, Sinisa STEKOVIC, Vincent Lepetit
Принадлежит: Qualcomm Inc

Systems and techniques are provided for determining environmental layouts. For example, based on one or more images of an environment and depth information associated with the one or more images, a set of candidate layouts and a set of candidate objects corresponding to the environment can be detected. The set of candidate layouts and set of candidate objects can be organized as a structured tree. For instance, a structured tree can be generated including nodes corresponding to the set of candidate layouts and the set of candidate objects. A combination of objects and layouts can be selected in the structured tree (e.g., based on a search of the structured tree, such as using a Monte-Carlo Tree Search (MCTS) algorithm or adapted MCTS algorithm). A three-dimensional (3D) layout of the environment can be determined based on the combination of objects and layouts in the structured tree.

Подробнее
Номер записи: 17
31-10-2023 дата публикации

Keypoint-based sampling for pose estimation

Номер: US11804040B2
Автор: Clemens Arth, Shreyas HAMPALI, Vincent Lepetit
Принадлежит: Qualcomm Inc

Systems and techniques are provided for determining one or more poses of one or more objects. For example, a process can include determining, using a machine learning system, a plurality of keypoints from an image. The plurality of keypoints are associated with at least one object in the image. The process can include determining a plurality of features from the machine learning system based on the plurality of keypoints. The process can include classifying the plurality of features into a plurality of joint types. The process can include determining pose parameters for the at least one object based on the plurality of joint types.

Подробнее
Номер записи: 18
16-01-2019 дата публикации

Systems and methods for normalizing an image

Номер: EP3427186A1
Автор: Mahdi RAD, Markus OBERWEGER, Vincent Lepetit
Принадлежит: Qualcomm Inc

A method for normalizing an image by an electronic device is described. The method includes obtaining an image including a target object. The method also includes determining a set of windows of the image. The method further includes, for each window of the set of windows of the image, predicting parameters of an illumination normalization model adapted to the window using a first convolutional neural network (CNN), and applying the illumination normalization model to the window to produce a normalized window.

Подробнее
Номер записи: 19
24-01-2024 дата публикации

Keypoint-based sampling for pose estimation

Номер: EP4309151A1
Автор: Clemens Arth, Shreyas HAMPALI, Vincent Lepetit
Принадлежит: Qualcomm Technologies Inc

Systems and techniques are provided for determining one or more poses of one or more objects. For example, a process can include determining, using a machine learning system, a plurality of keypoints from an image. The plurality of keypoints are associated with at least one object in the image. The process can include determining a plurality of features from the machine learning system based on the plurality of keypoints. The process can include classifying the plurality of features into a plurality of joint types. The process can include determining pose parameters for the at least one object based on the plurality of joint types.

Подробнее
Номер записи: 20
15-02-2024 дата публикации

Systems and methods for grasping objects like humans using robot grippers

Номер: US20240051125A1
Автор: Philippe Weinzaepfel, Romain Bregier, Vincent Lepetit, Yuming Du
Принадлежит: Ecole Nationale des Ponts et Chaussess ENPC, Naver Corp, Naver Labs Corp

A system includes: a hand module to, based on a demonstration of a human hand grasping an object, determine first and second vectors that are normal to and parallel to a palm of the human hand, respectively, and a position of the human hand; a gripper module to determine third and fourth vectors that are normal to and parallel to a palm of a gripper of a robot, respectively, and a present position of the gripper; and an actuation module to: move the gripper when open such that the present position of the gripper is at the position of the human hand, the third and first vectors are aligned, and the fourth and second vectors are aligned; close fingers of the gripper based on minimizing a first loss; and actuate the fingers of the gripper to minimize a second loss determined based on the first loss and a third loss.

Подробнее
Номер записи: 21
12-06-2018 дата публикации

Predictor-corrector based pose detection

Номер: US09996936B2
Автор: Clemens Arth, Paul WOHLHART, Vincent Lepetit
Принадлежит: Qualcomm Inc

A computer-implemented method, apparatus, computer readable medium and mobile device for determining a 6DOF pose from an input image. The process of determining 6DOF pose may include processing an input image to create one or more static representations of the input image, creating a dynamic representation of the input image from an estimated 6DOF pose and a 2.5D reference map, and measuring correlation between the dynamic representation and the one or more static representations of the input image. The estimated 6DOF may be iteratively adjusted according to the measured correlation error until a final adjusted dynamic representation meets an output threshold.

Подробнее
Номер записи: 22