Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Bidirectional Siamese Recurrent Neural Network for Accurate Gait Recognition Using Body Landmarks

Proma Hossain Progga, Md. Jobayer Rahman, Swapnil Biswas, Md. Shakil Ahmed, Arif Reza Anwary, Swakkhar Shatabda

TL;DR

The paper tackles cross-view gait recognition by combining Mediapipe-derived sequential landmarks with Procrustes alignment and a Siamese biGRU-dualStack network trained with contrastive loss. It demonstrates strong, consistent performance across CASIA-B, SZU RGB-D, OU-MVLP, and Gait3D, reporting high Rank-1 accuracies and favorable ablation results that justify bidirectional stacked temporal modeling. Key contributions include the landmark-based gait representation, view-normalizing alignment, a dual-branch Siamese architecture, and extensive empirical validation with diverse datasets and analysis of landmark subsets and clothing effects. The work offers a practical, efficient pathway for accurate gait-based person identification in surveillance and security contexts, with code forthcoming at the project repository.

Abstract

Gait recognition is a significant biometric technique for person identification, particularly in scenarios where other physiological biometrics are impractical or ineffective. In this paper, we address the challenges associated with gait recognition and present a novel approach to improve its accuracy and reliability. The proposed method leverages advanced techniques, including sequential gait landmarks obtained through the Mediapipe pose estimation model, Procrustes analysis for alignment, and a Siamese biGRU-dualStack Neural Network architecture for capturing temporal dependencies. Extensive experiments were conducted on large-scale cross-view datasets to demonstrate the effectiveness of the approach, achieving high recognition accuracy compared to other models. The model demonstrated accuracies of 95.7%, 94.44%, 87.71%, and 86.6% on CASIA-B, SZU RGB-D, OU-MVLP, and Gait3D datasets respectively. The results highlight the potential applications of the proposed method in various practical domains, indicating its significant contribution to the field of gait recognition.

A Bidirectional Siamese Recurrent Neural Network for Accurate Gait Recognition Using Body Landmarks

TL;DR

The paper tackles cross-view gait recognition by combining Mediapipe-derived sequential landmarks with Procrustes alignment and a Siamese biGRU-dualStack network trained with contrastive loss. It demonstrates strong, consistent performance across CASIA-B, SZU RGB-D, OU-MVLP, and Gait3D, reporting high Rank-1 accuracies and favorable ablation results that justify bidirectional stacked temporal modeling. Key contributions include the landmark-based gait representation, view-normalizing alignment, a dual-branch Siamese architecture, and extensive empirical validation with diverse datasets and analysis of landmark subsets and clothing effects. The work offers a practical, efficient pathway for accurate gait-based person identification in surveillance and security contexts, with code forthcoming at the project repository.

Abstract

Gait recognition is a significant biometric technique for person identification, particularly in scenarios where other physiological biometrics are impractical or ineffective. In this paper, we address the challenges associated with gait recognition and present a novel approach to improve its accuracy and reliability. The proposed method leverages advanced techniques, including sequential gait landmarks obtained through the Mediapipe pose estimation model, Procrustes analysis for alignment, and a Siamese biGRU-dualStack Neural Network architecture for capturing temporal dependencies. Extensive experiments were conducted on large-scale cross-view datasets to demonstrate the effectiveness of the approach, achieving high recognition accuracy compared to other models. The model demonstrated accuracies of 95.7%, 94.44%, 87.71%, and 86.6% on CASIA-B, SZU RGB-D, OU-MVLP, and Gait3D datasets respectively. The results highlight the potential applications of the proposed method in various practical domains, indicating its significant contribution to the field of gait recognition.

Paper Structure

This paper contains 14 sections, 12 equations, 8 figures, 8 tables.

Table of Contents

  1. Introduction
  2. Related work
  3. Methodology
  4. Experimental Analysis
  5. Dataset
  6. Data Pre-processing
  7. Procrustes Analysis
  8. Model Training
  9. Evaluation Metrics
  10. Individual Gait Differences
  11. Comparison with the SOTA Methods
  12. Ablation Study
  13. Discussion
  14. Conclusion

Figures (8)

  • Figure 1: Access Control Based on Gait Sequence Matching: Successful vs. Unsuccessful Cases.
  • Figure 2: Proposed Framework for Gait Recognition Using Sequential Landmarks.
  • Figure 3: Siamese BiGRU-dualStack Neural Network Architecture.
  • Figure 4: Overview of CASIA-B and SZU RGB-D Gait Datasets.
  • Figure 5: Landmark Shape Alignment Using Procrustes Analysis.
  • ...and 3 more figures