TrackletGait: A Robust Framework for Gait Recognition in the Wild

TL;DR

TrackletGait tackles gait recognition in the wild by addressing non-periodic, occluded silhouettes with a compact yet effective framework. It introduces Random Tracklet Sampling to capture diverse walking states, Haar Wavelet-based Downsampling to preserve spatial information, and Hardness Exclusion Triplet Loss to filter low-quality silhouettes during training. A 22-layer ResNet backbone with P3D-HWD units enables strong performance with only 10.3M parameters, achieving 77.8% rank-1 on Gait3D and 80.4% on GREW. Extensive ablations and cross-dataset analyses validate the approach as robust and efficient for real-world surveillance scenarios.

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.

Figures (8)

• Figure 1: Dataset collection scenarios: Gait recognition in the wild vs. gait recognition in the lab.
• Figure 2: Framework of the proposed TrackletGait.
• Figure 3: Illustration of three sampling methods.
• Figure 4: P3D-HWD residual unit.
• Figure 5: Illustration of the Haar Discrete Wavelet Transform.