TrackletGait: A Robust Framework for Gait Recognition in the Wild

2025-08-04

AI Summary

To address incomplete and distorted gait silhouette sequences caused by non-periodic walking patterns and severe occlusions in real-world surveillance scenarios, this paper proposes a robust end-to-end gait recognition framework. Methodologically: (1) random trajectory segment sampling is introduced to jointly enhance temporal robustness and discriminative representation learning; (2) Haar wavelet-based downsampling replaces conventional pooling to preserve critical spatial structural information while reducing resolution; (3) a hard-sample exclusion triplet loss is designed to automatically suppress interference from low-quality silhouettes. Evaluated on Gait3D and GREW benchmarks, the method achieves rank-1 accuracies of 77.8% and 80.4%, respectively—setting new state-of-the-art performance with only 10.3 million backbone parameters. This significantly improves identity recognition stability and accuracy under challenging surveillance conditions.

Technology Category

• Computer Vision: Biometrics, Face, Gesture & Pose
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Planning, Routing, and Scheduling: Activity and Plan Recognition

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Large-scale security measurements
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Gait recognition aims to identify individuals based on their body shape and walking patterns. Though much progress has been achieved driven by deep learning, gait recognition in real-world surveillance scenarios remains quite challenging to current methods. Conventional approaches, which rely on periodic gait cycles and controlled environments, struggle with the non-periodic and occluded silhouette sequences encountered in the wild. In this paper, we propose a novel framework, TrackletGait, designed to address these challenges in the wild. We propose Random Tracklet Sampling, a generalization of existing sampling methods, which strikes a balance between robustness and representation in capturing diverse walking patterns. Next, we introduce Haar Wavelet-based Downsampling to preserve information during spatial downsampling. Finally, we present a Hardness Exclusion Triplet Loss, designed to exclude low-quality silhouettes by discarding hard triplet samples. TrackletGait achieves state-of-the-art results, with 77.8 and 80.4 rank-1 accuracy on the Gait3D and GREW datasets, respectively, while using only 10.3M backbone parameters. Extensive experiments are also conducted to further investigate the factors affecting gait recognition in the wild.

Problem

• Address gait recognition challenges in real-world surveillance scenarios
• Handle non-periodic and occluded silhouette sequences in the wild
• Improve robustness and representation of diverse walking patterns

Innovation

• Random Tracklet Sampling for diverse walking patterns
• Haar Wavelet-based Downsampling preserves spatial information
• Hardness Exclusion Triplet Loss filters low-quality silhouettes