Multiperson Activity Recognition and Tracking Based on Skeletal Keypoint Detection | IEEE Journals & Magazine | IEEE Xplore

Multiperson Activity Recognition and Tracking Based on Skeletal Keypoint Detection


Impact Statement:Human Activity Recognition (HAR) technology is a hot research topic in the field of computer vision. However, most of the current action recognition networks require high...Show More

Abstract:

Currently, most action recognition networks have deep overall structures, large model parameters, and high requirements for computer hardware equipment. As a result, it i...Show More
Impact Statement:
Human Activity Recognition (HAR) technology is a hot research topic in the field of computer vision. However, most of the current action recognition networks require high computer hardware equipment and are difficult to detect in real time. By constructing an action recognition network that integrates human pose estimation and multi-target tracking, we solve the problem of too large parameters and difficulty in feature extraction due to video interference information. Our proposed method can effectively extract the keypoints of multi-person bones and track the target in the video. It improve that accuracy and speed of multi-person action tracking and recognition in complex scene.

Abstract:

Currently, most action recognition networks have deep overall structures, large model parameters, and high requirements for computer hardware equipment. As a result, it is easy to overfit in the recognition process for too deep network layers. Furthermore, it is also difficult to extract features because of the video's interference information, such as illumination and occlusion. To solve the above problems, we propose a multiperson action recognition and tracking algorithm based on skeletal keypoint detection. First, the n network combining the improved dense convolutional network and part affinity field is used to extract the skeletal information points of the human body. Then, we present an improved DeepSort network for multiperson target tracking, which contains a Hungarian matching algorithm based on the generalized intersection over union and a pedestrian reidentification network combining GhostNet and feature pyramid network. Finally, we construct a deep neural network model to ...
Published in: IEEE Transactions on Artificial Intelligence ( Volume: 5, Issue: 5, May 2024)
Page(s): 2279 - 2292
Date of Publication: 25 September 2023
Electronic ISSN: 2691-4581

Funding Agency:


References

References is not available for this document.