Causality-inspired Discriminative Feature Learning in Triple Domains for Gait Recognition

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Xiong, Haijun, Feng, Bin, Wang, Xinggang, Liu, Wenyu
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916328345960448
author Xiong, Haijun
Feng, Bin
Wang, Xinggang
Liu, Wenyu
author_facet Xiong, Haijun
Feng, Bin
Wang, Xinggang
Liu, Wenyu
contents Gait recognition is a biometric technology that distinguishes individuals by their walking patterns. However, previous methods face challenges when accurately extracting identity features because they often become entangled with non-identity clues. To address this challenge, we propose CLTD, a causality-inspired discriminative feature learning module designed to effectively eliminate the influence of confounders in triple domains, \ie, spatial, temporal, and spectral. Specifically, we utilize the Cross Pixel-wise Attention Generator (CPAG) to generate attention distributions for factual and counterfactual features in spatial and temporal domains. Then, we introduce the Fourier Projection Head (FPH) to project spatial features into the spectral space, which preserves essential information while reducing computational costs. Additionally, we employ an optimization method with contrastive learning to enforce semantic consistency constraints across sequences from the same subject. Our approach has demonstrated significant performance improvements on challenging datasets, proving its effectiveness. Moreover, it can be seamlessly integrated into existing gait recognition methods.
format Preprint
id arxiv_https___arxiv_org_abs_2407_12519
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Causality-inspired Discriminative Feature Learning in Triple Domains for Gait Recognition
Xiong, Haijun
Feng, Bin
Wang, Xinggang
Liu, Wenyu
Computer Vision and Pattern Recognition
Gait recognition is a biometric technology that distinguishes individuals by their walking patterns. However, previous methods face challenges when accurately extracting identity features because they often become entangled with non-identity clues. To address this challenge, we propose CLTD, a causality-inspired discriminative feature learning module designed to effectively eliminate the influence of confounders in triple domains, \ie, spatial, temporal, and spectral. Specifically, we utilize the Cross Pixel-wise Attention Generator (CPAG) to generate attention distributions for factual and counterfactual features in spatial and temporal domains. Then, we introduce the Fourier Projection Head (FPH) to project spatial features into the spectral space, which preserves essential information while reducing computational costs. Additionally, we employ an optimization method with contrastive learning to enforce semantic consistency constraints across sequences from the same subject. Our approach has demonstrated significant performance improvements on challenging datasets, proving its effectiveness. Moreover, it can be seamlessly integrated into existing gait recognition methods.
title Causality-inspired Discriminative Feature Learning in Triple Domains for Gait Recognition
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2407.12519