Bridging the Distribution Gap of Visible-Infrared Person Re-identification with Modality Batch Normalization

TL;DR

This paper introduces Modality Batch Normalization (MBN), a novel normalization technique that reduces distribution gaps in visible-infrared person re-identification, significantly improving model performance across various datasets and architectures.

Contribution

The paper proposes MBN, a new batch normalization method that normalizes each modality separately to address distribution gaps in VI-ReID tasks, enhancing model accuracy.

Findings

MBN effectively reduces distribution gaps in VI-ReID.

Models with MBN outperform baselines across datasets.

MBN improves performance with different backbones and loss functions.

Abstract

Visible-infrared cross-modality person re-identification (VI-ReID), whose aim is to match person images between visible and infrared modality, is a challenging cross-modality image retrieval task. Most existing works integrate batch normalization layers into their neural network, but we found out that batch normalization layers would lead to two types of distribution gap: 1) inter-mini-batch distribution gap -- the distribution gap of the same modality between each mini-batch; 2) intra-mini-batch modality distribution gap -- the distribution gap of different modality within the same mini-batch. To address these problems, we propose a new batch normalization layer called Modality Batch Normalization (MBN), which normalizes each modality sub-mini-batch respectively instead of the whole mini-batch, and can reduce these distribution gap significantly. Extensive experiments show that our MBN is able to boost the performance of VI-ReID models, even with different datasets, backbones and losses.