Structural-Semantic Reciprocal Learning for Unsupervised Visible-Infrared Person Re-Identification

2026-07-16

AI Summary

This work addresses the challenges of large modality discrepancy and the absence of cross-modal annotations in unsupervised visible-infrared person re-identification by proposing a structure–semantic mutual learning framework. The method leverages fine-grained structure disentanglement (FSD) to extract reliable body-part features as spatial anchors and introduces a closed-loop semantic calibration (CSC) mechanism that dynamically refines shared semantic prototypes, effectively mitigating the open-loop propagation of pseudo-label noise. By integrating self-correcting closed-loop alignment with clustering strategies, the framework enables robust cross-modal representation learning. It achieves state-of-the-art performance on both SYSU-MM01 and RegDB benchmarks, even surpassing several supervised methods on RegDB.

Abstract

Unsupervised visible-infrared person re-identification (USVI-ReID) is challenging due to the large modality gap and the lack of cross-modal identity annotations. Progressive association paradigms have been proposed to gradually bridge the gap, but they suffer from two critical bottlenecks: reliance on ambiguous global representations and unchecked propagation of pseudo-label noise in an open-loop manner. To address these issues, we propose Structural-Semantic Reciprocal Learning (SSRL), a framework that transforms open-loop association into a self-correcting closed-loop system. Structurally, we introduce Fine-grained Structural Decoupling (FSD) to extract discriminative body-part primitives as reliable spatial anchors, complementing ambiguous holistic silhouettes with spatially consistent structural details. Semantically, we design a Closed-loop Semantic Calibration (CSC) mechanism that reconstructs shared semantic prototypes at each epoch and feeds them back into the training loop, effectively filtering pseudo-label noise before the next clustering cycle. Through the reciprocal interaction between structural and semantic learning, SSRL achieves robust cross-modal representation. Extensive experiments demonstrate the competitive performance of SSRL against state-of-the-art USVI-ReID methods on both SYSU-MM01 and RegDB, notably surpassing several supervised counterparts on RegDB.

Problem

• unsupervised visible-infrared person re-identification
• modality gap
• pseudo-label noise
• cross-modal identity annotations

Innovation

• Structural-Semantic Reciprocal Learning
• Fine-grained Structural Decoupling
• Closed-loop Semantic Calibration
• Unsupervised Visible-Infrared ReID
• Pseudo-label Noise Filtering