Learning to Balance: Diverse Normalization for Cloth-Changing Person Re-Identification

TL;DR

This paper introduces Diverse Norm, a novel normalization module that enhances cloth-changing person re-identification by effectively separating clothing and identity features without extra data, outperforming existing methods.

Contribution

The paper proposes Diverse Norm, a new normalization technique with orthogonal feature expansion and channel attention, improving cloth-changing person re-ID without auxiliary data.

Findings

Diverse Norm significantly outperforms state-of-the-art methods.

The approach effectively separates clothing and identity features.

No additional data is required for the method.

Abstract

Cloth-Changing Person Re-Identification (CC-ReID) involves recognizing individuals in images regardless of clothing status. In this paper, we empirically and experimentally demonstrate that completely eliminating or fully retaining clothing features is detrimental to the task. Existing work, either relying on clothing labels, silhouettes, or other auxiliary data, fundamentally aim to balance the learning of clothing and identity features. However, we practically find that achieving this balance is challenging and nuanced. In this study, we introduce a novel module called Diverse Norm, which expands personal features into orthogonal spaces and employs channel attention to separate clothing and identity features. A sample re-weighting optimization strategy is also introduced to guarantee the opposite optimization direction. Diverse Norm presents a simple yet effective approach that does not require additional data. Furthermore, Diverse Norm can be seamlessly integrated ResNet50 and significantly outperforms the state-of-the-art methods.