Dual-level Modality Debiasing Learning for Unsupervised Visible-Infrared Person Re-Identification

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Jiaze, Lu, Yan, Liu, Bin, Yin, Guojun, Ye, Mang
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917393665622016
author Li, Jiaze
Lu, Yan
Liu, Bin
Yin, Guojun
Ye, Mang
author_facet Li, Jiaze
Lu, Yan
Liu, Bin
Yin, Guojun
Ye, Mang
contents Two-stage learning pipeline has achieved promising results in unsupervised visible-infrared person re-identification (USL-VI-ReID). It first performs single-modality learning and then operates cross-modality learning to tackle the modality discrepancy. Although promising, this pipeline inevitably introduces modality bias: modality-specific cues learned in the single-modality training naturally propagate into the following cross-modality learning, impairing identity discrimination and generalization. To address this issue, we propose a Dual-level Modality Debiasing Learning (DMDL) framework that implements debiasing at both the model and optimization levels. At the model level, we propose a Causality-inspired Adjustment Intervention (CAI) module that replaces likelihood-based modeling with causal modeling, preventing modality-induced spurious patterns from being introduced, leading to a low-biased model. At the optimization level, a Collaborative Bias-free Training (CBT) strategy is introduced to interrupt the propagation of modality bias across data, labels, and features by integrating modality-specific augmentation, label refinement, and feature alignment. Extensive experiments on benchmark datasets demonstrate that DMDL could enable modality-invariant feature learning and a more generalized model. The code is available at https://github.com/priester3/DMDL.
format Preprint
id arxiv_https___arxiv_org_abs_2512_03745
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Dual-level Modality Debiasing Learning for Unsupervised Visible-Infrared Person Re-Identification
Li, Jiaze
Lu, Yan
Liu, Bin
Yin, Guojun
Ye, Mang
Computer Vision and Pattern Recognition
Two-stage learning pipeline has achieved promising results in unsupervised visible-infrared person re-identification (USL-VI-ReID). It first performs single-modality learning and then operates cross-modality learning to tackle the modality discrepancy. Although promising, this pipeline inevitably introduces modality bias: modality-specific cues learned in the single-modality training naturally propagate into the following cross-modality learning, impairing identity discrimination and generalization. To address this issue, we propose a Dual-level Modality Debiasing Learning (DMDL) framework that implements debiasing at both the model and optimization levels. At the model level, we propose a Causality-inspired Adjustment Intervention (CAI) module that replaces likelihood-based modeling with causal modeling, preventing modality-induced spurious patterns from being introduced, leading to a low-biased model. At the optimization level, a Collaborative Bias-free Training (CBT) strategy is introduced to interrupt the propagation of modality bias across data, labels, and features by integrating modality-specific augmentation, label refinement, and feature alignment. Extensive experiments on benchmark datasets demonstrate that DMDL could enable modality-invariant feature learning and a more generalized model. The code is available at https://github.com/priester3/DMDL.
title Dual-level Modality Debiasing Learning for Unsupervised Visible-Infrared Person Re-Identification
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.03745