-
公开(公告)号：US20190251397A1
公开(公告)日：2019-08-15
申请号：US16256820
申请日：2019-01-24
Applicant: NVIDIA Corporation
Inventor： Jonathan Tremblay , Aayush Prakash , Mark A. Brophy , Varun Jampani , Cem Anil , Stanley Thomas Birchfield , Thang Hong To , David Jesus Acuna Marrero
CPC classification number: G06K9/6257 , G06K9/6228 , G06K2209/21 , G06N3/04 , G06N3/084 , G06T15/04 , G06T15/20 , G06T15/50 , G06T2210/12
Abstract: Training deep neural networks requires a large amount of labeled training data. Conventionally, labeled training data is generated by gathering real images that are manually labelled which is very time-consuming. Instead of manually labelling a training dataset, domain randomization technique is used generate training data that is automatically labeled. The generated training data may be used to train neural networks for object detection and segmentation (labelling) tasks. In an embodiment, the generated training data includes synthetic input images generated by rendering three-dimensional (3D) objects of interest in a 3D scene. In an embodiment, the generated training data includes synthetic input images generated by rendering 3D objects of interest on a 2D background image. The 3D objects of interest are objects that a neural network is trained to detect and/or label.
-
公开(公告)号：US12194632B2
公开(公告)日：2025-01-14
申请号：US18378241
申请日：2023-10-10
Applicant: NVIDIA Corporation
Inventor： Shariq Iqbal , Jonathan Tremblay , Thang Hong To , Jia Cheng , Erik Leitch , Duncan J. McKay , Stanley Thomas Birchfield
Abstract: In at least one embodiment, under the control of a robotic control system, a gripper on a robot is positioned to grasp a 3-dimensional object. In at least one embodiment, the relative position of the object and the gripper is determined, at least in part, by using a camera mounted on the gripper.
-
公开(公告)号：US11715251B2
公开(公告)日：2023-08-01
申请号：US17507620
申请日：2021-10-21
Applicant: NVIDIA Corporation
Inventor： Jonathan Tremblay , Aayush Prakash , Mark A. Brophy , Varun Jampani , Cem Anil , Stanley Thomas Birchfield , Thang Hong To , David Jesus Acuna Marrero
IPC: G06T15/00 , G06T15/04 , G06T15/50 , G06T15/20 , G06F18/214 , G06F18/211 , G06V10/774 , G06V10/82 , G06N3/04 , G06N3/084
CPC classification number: G06T15/00 , G06F18/211 , G06F18/2148 , G06T15/04 , G06T15/20 , G06T15/50 , G06V10/7747 , G06V10/82 , G06N3/04 , G06N3/084 , G06T2210/12 , G06V2201/07
Abstract: Training deep neural networks requires a large amount of labeled training data. Conventionally, labeled training data is generated by gathering real images that are manually labelled which is very time-consuming. Instead of manually labelling a training dataset, domain randomization technique is used generate training data that is automatically labeled. The generated training data may be used to train neural networks for object detection and segmentation (labelling) tasks. In an embodiment, the generated training data includes synthetic input images generated by rendering three-dimensional (3D) objects of interest in a 3D scene. In an embodiment, the generated training data includes synthetic input images generated by rendering 3D objects of interest on a 2D background image. The 3D objects of interest are objects that a neural network is trained to detect and/or label.
-
公开(公告)号：US11074717B2
公开(公告)日：2021-07-27
申请号：US16405662
申请日：2019-05-07
Applicant: NVIDIA Corporation
Inventor： Jonathan Tremblay , Thang Hong To , Stanley Thomas Birchfield
Abstract: An object detection neural network receives an input image including an object and generates belief maps for vertices of a bounding volume that encloses the object. The belief maps are used, along with three-dimensional (3D) coordinates defining the bounding volume, to compute the pose of the object in 3D space during post-processing. When multiple objects are present in the image, the object detection neural network may also generate vector fields for the vertices. A vector field comprises vectors pointing from the vertex to a centroid of the object enclosed by the bounding volume defined by the vertex. The object detection neural network may be trained using images of computer-generated objects rendered in 3D scenes (e.g., photorealistic synthetic data). Automatically labelled training datasets may be easily constructed using the photorealistic synthetic data. The object detection neural network may be trained for object detection using only the photorealistic synthetic data.
-
公开(公告)号：US20210097346A1
公开(公告)日：2021-04-01
申请号：US17119971
申请日：2020-12-11
Applicant: NVIDIA Corporation
Inventor： Jonathan Tremblay , Aayush Prakash , Mark A. Brophy , Varun Jampani , Cem Anil , Stanley Thomas Birchfield , Thang Hong To , David Jesus Acuna Marrero
Abstract: Training deep neural networks requires a large amount of labeled training data. Conventionally, labeled training data is generated by gathering real images that are manually labelled which is very time-consuming. Instead of manually labelling a training dataset, domain randomization technique is used generate training data that is automatically labeled. The generated training data may be used to train neural networks for object detection and segmentation (labelling) tasks. In an embodiment, the generated training data includes synthetic input images generated by rendering three-dimensional (3D) objects of interest in a 3D scene. In an embodiment, the generated training data includes synthetic input images generated by rendering 3D objects of interest on a 2D background image. The 3D objects of interest are objects that a neural network is trained to detect and/or label.
-
Patent Agency Ranking
公开(公告)号：US20200061811A1
公开(公告)日：2020-02-27
申请号：US16549831
申请日：2019-08-23
Applicant: NVIDIA Corporation
Inventor： Shariq Iqbal , Jonathan Tremblay , Thang Hong To , Jia Cheng , Erik Leitch , Duncan J. McKay , Stanley Thomas Birchfield
Abstract: In at least one embodiment, under the control of a robotic control system, a gripper on a robot is positioned to grasp a 3-dimensional object. In at least one embodiment, the relative position of the object and the gripper is determined, at least in part, by using a camera mounted on the gripper.
-
公开(公告)号：US20190355150A1
公开(公告)日：2019-11-21
申请号：US16405662
申请日：2019-05-07
Applicant: NVIDIA Corporation
Inventor： Jonathan Tremblay , Thang Hong To , Stanley Thomas Birchfield
Abstract: An object detection neural network receives an input image including an object and generates belief maps for vertices of a bounding volume that encloses the object. The belief maps are used, along with three-dimensional (3D) coordinates defining the bounding volume, to compute the pose of the object in 3D space during post-processing. When multiple objects are present in the image, the object detection neural network may also generate vector fields for the vertices. A vector field comprises vectors pointing from the vertex to a centroid of the object enclosed by the bounding volume defined by the vertex. The object detection neural network may be trained using images of computer-generated objects rendered in 3D scenes (e.g., photorealistic synthetic data). Automatically labelled training datasets may be easily constructed using the photorealistic synthetic data. The object detection neural network may be trained for object detection using only the photorealistic synthetic data.
-
公开(公告)号：US20240042601A1
公开(公告)日：2024-02-08
申请号：US18378241
申请日：2023-10-10
Applicant: NVIDIA Corporation
Inventor： Shariq Iqbal , Jonathan Tremblay , Thang Hong To , Jia Cheng , Erik Leitch , Duncan J. McKay , Stanley Thomas Birchfield
CPC classification number: B25J9/161 , B25J9/1612 , G06T7/74 , G06N3/08 , B25J9/1669 , G06T2207/30244
Abstract: In at least one embodiment, under the control of a robotic control system, a gripper on a robot is positioned to grasp a 3-dimensional object. In at least one embodiment, the relative position of the object and the gripper is determined, at least in part, by using a camera mounted on the gripper.
-
公开(公告)号：US11833681B2
公开(公告)日：2023-12-05
申请号：US16549831
申请日：2019-08-23
Applicant: NVIDIA Corporation
Inventor： Shariq Iqbal , Jonathan Tremblay , Thang Hong To , Jia Cheng , Erik Leitch , Duncan J. McKay , Stanley Thomas Birchfield
CPC classification number: B25J9/161 , B25J9/1612 , B25J9/1669 , G06N3/08 , G06T7/74 , G06T2207/30244
Abstract: In at least one embodiment, under the control of a robotic control system, a gripper on a robot is positioned to grasp a 3-dimensional object. In at least one embodiment, the relative position of the object and the gripper is determined, at least in part, by using a camera mounted on the gripper.
-
公开(公告)号：US11182649B2
公开(公告)日：2021-11-23
申请号：US17119971
申请日：2020-12-11
Applicant: NVIDIA Corporation
Inventor： Jonathan Tremblay , Aayush Prakash , Mark A. Brophy , Varun Jampani , Cem Anil , Stanley Thomas Birchfield , Thang Hong To , David Jesus Acuna Marrero
Abstract: Training deep neural networks requires a large amount of labeled training data. Conventionally, labeled training data is generated by gathering real images that are manually labelled which is very time-consuming. Instead of manually labelling a training dataset, domain randomization technique is used generate training data that is automatically labeled. The generated training data may be used to train neural networks for object detection and segmentation (labelling) tasks. In an embodiment, the generated training data includes synthetic input images generated by rendering three-dimensional (3D) objects of interest in a 3D scene. In an embodiment, the generated training data includes synthetic input images generated by rendering 3D objects of interest on a 2D background image. The 3D objects of interest are objects that a neural network is trained to detect and/or label.
-
-
-
-
-
-
-
-
-
Search Results
10
records
(took 0.015 seconds)
