A robust event-driven gait recognition method, system, device and storage medium based on event stream

By dividing the event stream into time slices and normalizing features, and combining graph convolutional networks and loss function optimization, the problems of graph structure robustness and instability of deep message passing in event gait recognition are solved, thereby improving the robustness and accuracy of recognition.

CN122176608APending Publication Date: 2026-06-09CHINA UNIV OF PETROLEUM (EAST CHINA)
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN Β· China
Patent Type
Applications(China)
Current Assignee / Owner
CHINA UNIV OF PETROLEUM (EAST CHINA)
Filing Date
2026-03-19
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Existing event gait recognition methods based on graph neural networks have significant shortcomings in the robustness of graph structure construction, the stability of deep message passing, and the discriminative power of fine-grained features.

Method used

The denoised event stream is divided into continuous time slices and an event graph is constructed. Local and global time cues are added, spatial displacement, temporal displacement and edge length are renormalized, motion intensity index is introduced to determine edge reliability, graph convolutional backbone network is used to extract features, and temporal relationships are jointly modeled by difference branches and similar branches to optimize additive angular interval loss and dynamic center loss.

Benefits of technology

It significantly alleviates statistical drift caused by pooling, improves the stability and feature discrimination of deep graph message passing, and enhances the robustness and accuracy of gait recognition, making it suitable for complex environments such as cross-viewpoint, low light, and speed variations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122176608A_ABST
    Figure CN122176608A_ABST
Patent Text Reader

Abstract

This invention discloses a robust event-driven gait recognition method, system, device, and storage medium based on event flow, relating to the fields of event vision, computer vision, pattern recognition, and intelligent security technologies. The method includes renormalizing spatial displacement, temporal displacement, and edge length according to a unified scale and robustness scale, recalculating edge attributes after each pooling, introducing a motion intensity index to assist in determining edge reliability, employing continuous reweighting instead of direct edge deletion, and using motion consistency, radius validity, orientation validity, and entropy constraints to weakly supervise edge confidence. A graph convolutional backbone network is used to extract spatial graph features for each time slice, and temporal relationships are jointly modeled through difference and similarity branches. The additive angular interval loss and dynamic center loss are jointly optimized. This invention improves message passing stability, enhances anti-disturbance robustness, overcomes intra-class fluctuations such as cross-viewpoint and low-light conditions, and possesses excellent potential for edge device deployment.
Need to check novelty before this filing date? Find Prior Art