Deep learning gait recognition based on two branch spatiotemporal gait feature fusion

Aiming at the problem that the existing gait recognition methods are easily affected by shooting angle and clothing changes, this paper proposes a deep learning gait recognition method that fuses 2D shoulderless pose topological energy maps (SPTEMs) and 3D local skeleton gait features (LSGFs). Firstly, the lightweight BlazePose pose estimation algorithm is used to extract the human posture topology in the gait video sequence to generate the SPTEM, which improves the detection speed and reduces the impact of clothing changes. Then, the LSGF is introduced to make up for the low recognition accuracy deficiency of a single energy map feature in the case of variable viewing angles. Finally, a spatiotemporal feature extraction network model fused with an attention mechanism is proposed, and the two-stream features are fused uniformly in the fully connected layer. The proposed algorithm is validated on the CASIA-B dataset and compared with the current mainstream gait recognition methods. The results show that the gait recognition rate of the proposed method is significantly improved under cross-view and cl conditions. © 2024 Northeast University. All rights reserved.