公开(公告)号：US20230386244A1

公开(公告)日：2023-11-30

申请号：US18078027

申请日：2022-12-08

Applicant: UBTECH ROBOTICS CORP LTD

Inventor： SHUPING HU ,  Kan Wang ,  Huan Tan ,  Jianxin Pang

IPC: G06V40/10 ,  G06V10/776 ,  G06V10/74

CPC classification number: G06V40/103 ,  G06V10/776 ,  G06V10/761

Abstract: A person re-identification method, a storage medium, and a terminal device are provided. In the method, a preset ratio-based triplet loss function is used as a loss function during training The ratio-based triplet loss function limits a ratio of a positive sample feature distance to a negative sample feature distance to be less than a preset ratio threshold. The positive sample feature distance is a distance between a reference image feature and a positive sample image feature, and the negative sample feature distance is a distance between the reference image feature and a negative sample image feature. Compared with the existing absolute distance-based triplet loss function, in the case of small inter-class differences and large intra-class differences, the ratio-based triplet loss function can effectively improve the stability of model training, the features extracted by the trained model are more discriminative and robust, thereby improving the accuracy of person re-identification results.

公开(公告)号：US20240290096A1

公开(公告)日：2024-08-29

申请号：US18422046

申请日：2024-01-25

Applicant: UBTECH ROBOTICS CORP LTD

Inventor： Kan Wang ,  Shuping Hu ,  Jianxin Pang ,  Huan Tan

IPC: G06V20/40 ,  G06V10/77 ,  G06V10/80

CPC classification number: G06V20/46 ,  G06V10/7715 ,  G06V10/806

Abstract: A method for extracting video frame features includes: obtaining a number of initial features of each video frame in a video sequence; calculating global channel attention information of the video sequence based on the initial features of each video frame in the video sequence; calculating local channel attention information of a target video frame according to initial features of a target video frame; wherein the target video frame is one of the video frames in the video sequence; and performing channel attention mechanism processing on the initial features of the target video frame according to the global channel attention information and the local channel attention information to obtain optimized features of the target video frame.

公开(公告)号：US20240290097A1

公开(公告)日：2024-08-29

申请号：US18427775

申请日：2024-01-30

Applicant: UBTech Robotics Corp Ltd

Inventor： Kan Wang ,  Jianxin Pang ,  Huan Tan

IPC: G06V20/40 ,  G06V10/74 ,  G06V10/77 ,  G06V10/82

CPC classification number: G06V20/46 ,  G06V10/761 ,  G06V10/7715 ,  G06V10/82 ,  G06V20/48

Abstract: A method for extracting video features may include: obtaining a target video sequence that comprises a number of video frames; performing video frame feature extraction on the target video sequence to obtain video frame features of each of the video frames; performing feature weight calculation on each of the video frame features to obtain the feature weight of each of the video frame features; wherein the feature weight of each of the video frame features is determined by the video frame features of all of the video frames in the target video sequence; and performing feature weighting on each of the video frame features according to the feature weight of each of the video frame features to obtain video features of the target video sequence.

公开(公告)号：US20240221345A1

公开(公告)日：2024-07-04

申请号：US18533199

申请日：2023-12-08

Applicant: UBTECH ROBOTICS CORP LTD

Inventor： Kan Wang ,  Jiaxin Pang ,  Huan Tan

IPC: G06V10/42 ,  G06V10/77 ,  G06V10/774 ,  G06V10/82 ,  G06V40/10

CPC classification number: G06V10/42 ,  G06V10/7715 ,  G06V10/774 ,  G06V10/82 ,  G06V40/10

Abstract: A method for pedestrian body part feature extraction is provided, including: performing global feature extraction on a target pedestrian image to obtain a global feature map; learning each of body parts in the global feature map using a self-produced supervision signals-based self-regulated channel attention model to output first channel attention vectors each describing a respective one of the body parts; weighting the first channel attention vectors with the global feature map to obtain a weighted feature map describing the body parts; and extracting body part features of the target pedestrian image from the weighted feature map.