BiggerGait: Unlocking Gait Recognition with Layer-wise Representations from Large Vision Models

TL;DR

This paper introduces BiggerGait, a simple universal baseline leveraging layer-wise representations from large vision models to improve gait recognition, demonstrating superior performance across multiple datasets and tasks.

Contribution

It reveals the importance of intermediate layer representations in large vision models for gait recognition and proposes a practical baseline method, BiggerGait, that outperforms existing approaches.

Findings

Layer-wise features from LVMs enhance gait recognition performance.

BiggerGait achieves superior results on multiple datasets.

The approach is effective across within- and cross-domain tasks.

Abstract

Large vision models (LVM) based gait recognition has achieved impressive performance. However, existing LVM-based approaches may overemphasize gait priors while neglecting the intrinsic value of LVM itself, particularly the rich, distinct representations across its multi-layers. To adequately unlock LVM's potential, this work investigates the impact of layer-wise representations on downstream recognition tasks. Our analysis reveals that LVM's intermediate layers offer complementary properties across tasks, integrating them yields an impressive improvement even without rich well-designed gait priors. Building on this insight, we propose a simple and universal baseline for LVM-based gait recognition, termed BiggerGait. Comprehensive evaluations on CCPG, CAISA-B*, SUSTech1K, and CCGR\_MINI validate the superiority of BiggerGait across both within- and cross-domain tasks, establishing it as a simple yet practical baseline for gait representation learning. All the models and code will be publicly available.