公开(公告)号：US20240312187A1

公开(公告)日：2024-09-19

申请号：US18184071

申请日：2023-03-15

Applicant: NVIDIA Corporation

Inventor： Yue Wu ,  Cheng-Chieh Yang ,  Xin Tong ,  Minwoo Park

IPC: G06V10/771 ,  G06V10/77

CPC classification number: G06V10/771 ,  G06V10/7715

Abstract: In various examples, feature tracking for autonomous or semi-autonomous systems and applications is described herein. Systems and methods are disclosed that merge, using one or more processes, features detected using a feature tracker(s) and features detected using a feature detector(s) in order to track features between images. In some examples, the number of merged features and/or the locations of the merged features within the images are limited. This way, the systems and methods are able to identify merged features that are of greater importance for tracking while refraining from tracking merged features that are of less importance. For example, if the systems and methods are being used to identify features for autonomous driving, a greater number of merged features that are associated with objects located proximate to the driving surface may be tracked as compared to merged features that are associated with the sky.

公开(公告)号：US20240062657A1

公开(公告)日：2024-02-22

申请号：US18491492

申请日：2023-10-20

Applicant: NVIDIA Corporation

Inventor： Yue Wu ,  Pekka Janis ,  Xin Tong ,  Cheng-Chieh Yang ,  Minwoo Park ,  David Nister

IPC: G08G1/16 ,  G06V10/82 ,  G06V20/58 ,  G06V20/10 ,  G06F18/214 ,  G05D1/00 ,  G05D1/02 ,  G06N3/04 ,  G06T7/20

CPC classification number: G08G1/166 ,  G06V10/82 ,  G06V20/58 ,  G06V20/10 ,  G06F18/214 ,  G05D1/0088 ,  G05D1/0289 ,  G06N3/0418 ,  G06T7/20 ,  G05D2201/0213

Abstract: In various examples, a sequential deep neural network (DNN) may be trained using ground truth data generated by correlating (e.g., by cross-sensor fusion) sensor data with image data representative of a sequences of images. In deployment, the sequential DNN may leverage the sensor correlation to compute various predictions using image data alone. The predictions may include velocities, in world space, of objects in fields of view of an ego-vehicle, current and future locations of the objects in image space, and/or a time-to-collision (TTC) between the objects and the ego-vehicle. These predictions may be used as part of a perception system for understanding and reacting to a current physical environment of the ego-vehicle.

公开(公告)号：US20220188608A1

公开(公告)日：2022-06-16

申请号：US17122598

申请日：2020-12-15

Applicant: NVIDIA Corporation

Inventor： Gregory Heinrich ,  Maxim Milakov ,  Xin Tong ,  Yue Wu

IPC: G06N3/063 ,  G06N5/04 ,  G06F12/0893

Abstract: Apparatuses, systems, and techniques to cache and reuse data for a neural network. In at least one embodiment, data generated by one or more layers of a neural network is cached and reused by the neural network.

公开(公告)号：US11579629B2

公开(公告)日：2023-02-14

申请号：US16514404

申请日：2019-07-17

Applicant: NVIDIA Corporation

Inventor： Yue Wu ,  Pekka Janis ,  Xin Tong ,  Cheng-Chieh Yang ,  Minwoo Park ,  David Nister

IPC: G05D1/02 ,  G06K9/62 ,  G06T7/20 ,  G05D1/00 ,  G06V20/10 ,  G06V20/58 ,  G06N3/04

Abstract: In various examples, a sequential deep neural network (DNN) may be trained using ground truth data generated by correlating (e.g., by cross-sensor fusion) sensor data with image data representative of a sequences of images. In deployment, the sequential DNN may leverage the sensor correlation to compute various predictions using image data alone. The predictions may include velocities, in world space, of objects in fields of view of an ego-vehicle, current and future locations of the objects in image space, and/or a time-to-collision (TTC) between the objects and the ego-vehicle. These predictions may be used as part of a perception system for understanding and reacting to a current physical environment of the ego-vehicle.

公开(公告)号：US20230122119A1

公开(公告)日：2023-04-20

申请号：US18067176

申请日：2022-12-16

Applicant: NVIDIA Corporation

Inventor： Yue Wu ,  Pekka Janis ,  Xin Tong ,  Cheng-Chieh Yang ,  Minwoo Park ,  David Nister

IPC: G05D1/02 ,  G06N3/04 ,  G06T7/20 ,  G05D1/00 ,  G06V20/10 ,  G06V20/58 ,  G06F18/214

Abstract: In various examples, a sequential deep neural network (DNN) may be trained using ground truth data generated by correlating (e.g., by cross-sensor fusion) sensor data with image data representative of a sequences of images. In deployment, the sequential DNN may leverage the sensor correlation to compute various predictions using image data alone. The predictions may include velocities, in world space, of objects in fields of view of an ego-vehicle, current and future locations of the objects in image space, and/or a time-to-collision (TTC) between the objects and the ego-vehicle. These predictions may be used as part of a perception system for understanding and reacting to a current physical environment of the ego-vehicle.

公开(公告)号：US11854401B2

公开(公告)日：2023-12-26

申请号：US18067176

申请日：2022-12-16

Applicant: NVIDIA Corporation

Inventor： Yue Wu ,  Pekka Janis ,  Xin Tong ,  Cheng-Chieh Yang ,  Minwoo Park ,  David Nister

IPC: G08G1/16 ,  G06V10/82 ,  G06V20/58 ,  G06V20/10 ,  G06F18/214 ,  G05D1/00 ,  G05D1/02 ,  G06N3/04 ,  G06T7/20

CPC classification number: G08G1/166 ,  G05D1/0088 ,  G05D1/0289 ,  G06F18/214 ,  G06N3/0418 ,  G06T7/20 ,  G06V10/82 ,  G06V20/10 ,  G06V20/58 ,  G05D2201/0213

Abstract: In various examples, a sequential deep neural network (DNN) may be trained using ground truth data generated by correlating (e.g., by cross-sensor fusion) sensor data with image data representative of a sequences of images. In deployment, the sequential DNN may leverage the sensor correlation to compute various predictions using image data alone. The predictions may include velocities, in world space, of objects in fields of view of an ego-vehicle, current and future locations of the objects in image space, and/or a time-to-collision (TTC) between the objects and the ego-vehicle. These predictions may be used as part of a perception system for understanding and reacting to a current physical environment of the ego-vehicle.

公开(公告)号：US20230351769A1

公开(公告)日：2023-11-02

申请号：US17733508

申请日：2022-04-29

Applicant: NVIDIA Corporation

Inventor： Yue WU ,  Liwen Lin ,  Xin Tong ,  Gang Pan

IPC: G06V20/58 ,  G06T7/55 ,  G06V10/74 ,  G06V10/764 ,  G06V10/762

CPC classification number: G06V20/58 ,  G06T7/55 ,  G06V10/761 ,  G06V10/764 ,  G06V10/762 ,  G06T2207/20081 ,  G06T2207/30261 ,  B60W60/0015

Abstract: In various examples, systems and methods for machine learning based hazard detection for autonomous machine applications using stereo disparity are presented. Disparity between a stereo pair of images is used to generate a path disparity model. Using the path disparity model, a machine learning model can recognize when a pixel in the first image corresponds to a pixel in the second image even though the pixel in the two images does not have identical characteristics. Similarities in extracted feature vectors can be computed and represented by a vector similarity metric that is input to a machine learning classifier, along with feature information extracted from the stereo image pair, to differentiate hazard pixels from non-hazard pixels. In some embodiments, a V-space disparity map, where a first axis corresponds to disparity values and the second axis corresponds to pixel rows, may be used to simplify estimation of the path disparity model.

公开(公告)号：US20200293064A1

公开(公告)日：2020-09-17

申请号：US16514404

申请日：2019-07-17

Applicant: NVIDIA Corporation

Inventor： Yue Wu ,  Pekka Janis ,  Xin Tong ,  Cheng-Chieh Yang ,  Minwoo Park ,  David Nister

IPC: G05D1/02 ,  G06N3/04 ,  G06K9/62 ,  G06K9/00 ,  G06T7/20 ,  G05D1/00

Abstract: In various examples, a sequential deep neural network (DNN) may be trained using ground truth data generated by correlating (e.g., by cross-sensor fusion) sensor data with image data representative of a sequences of images. In deployment, the sequential DNN may leverage the sensor correlation to compute various predictions using image data alone. The predictions may include velocities, in world space, of objects in fields of view of an ego-vehicle, current and future locations of the objects in image space, and/or a time-to-collision (TTC) between the objects and the ego-vehicle. These predictions may be used as part of a perception system for understanding and reacting to a current physical environment of the ego-vehicle.