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

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Ye, Dingqiang, Fan, Chao, Huang, Zhanbo, Luo, Chengwen, Li, Jianqiang, Yu, Shiqi, Liu, Xiaoming
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918061021331456
author Ye, Dingqiang
Fan, Chao
Huang, Zhanbo
Luo, Chengwen
Li, Jianqiang
Yu, Shiqi
Liu, Xiaoming
author_facet Ye, Dingqiang
Fan, Chao
Huang, Zhanbo
Luo, Chengwen
Li, Jianqiang
Yu, Shiqi
Liu, Xiaoming
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2505_18132
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle BiggerGait: Unlocking Gait Recognition with Layer-wise Representations from Large Vision Models
Ye, Dingqiang
Fan, Chao
Huang, Zhanbo
Luo, Chengwen
Li, Jianqiang
Yu, Shiqi
Liu, Xiaoming
Computer Vision and Pattern Recognition
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.
title BiggerGait: Unlocking Gait Recognition with Layer-wise Representations from Large Vision Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2505.18132