Style Normalization and Restitution for Generalizable Person Re-identification

Abstract

Existing fully-supervised person re-identification (ReID) methods usually suffer from poor generalization capability caused by domain gaps. The key to solving this problem lies in filtering out identity-irrelevant interference and learning domain-invariant person representations. In this paper, we aim to design a generalizable person ReID framework which trains a model on source domains yet is able to generalize/perform well on target domains. To achieve this goal, we propose a simple yet effective Style Normalization and Restitution (SNR) module. Specifically, we filter out style variations (e.g., illumination, color contrast) by Instance Normalization (IN). However, such a process inevitably removes discriminative information. We propose to distill identity-relevant feature from the removed information and restitute it to the network to ensure high discrimination. For better disentanglement, we enforce a dual causal loss constraint in SNR to encourage the separation of identity-relevant features and identity-irrelevant features. Extensive experiments demonstrate the strong generalization capability of our framework. Our models empowered by the SNR modules significantly outperform the state-of-the-art domain generalization approaches on multiple widely-used person ReID benchmarks, and also show superiority on unsupervised domain adaptation.

Accepted by CVPR2020

References in corpus (13)

• Instance Normalization: The Missing Ingredient for Fast Stylization
• In Defense of the Triplet Loss for Person Re-Identification
• Domain Generalization via Invariant Feature Representation
• Revisiting Batch Normalization For Practical Domain Adaptation
• Generalizing Across Domains via Cross-Gradient Training
• FD-GAN: Pose-guided Feature Distilling GAN for Robust Person Re-identification
• Unlabeled Samples Generated by GAN Improve the Person Re-identification Baseline in vitro
• Cross-Domain Visual Matching via Generalized Similarity Measure and Feature Learning
• Re-ID done right: towards good practices for person re-identification
• Unsupervised Domain Adaptive Re-Identification: Theory and Practice
• Unsupervised Person Re-identification by Soft Multilabel Learning
• Frustratingly Easy Person Re-Identification: Generalizing Person Re-ID in Practice
• Fairest of Them All: Establishing a Strong Baseline for Cross-Domain Person ReID

Cited by in corpus (2)

• Feature Alignment and Restoration for Domain Generalization and Adaptation
• Style Normalization and Restitution for Domain Generalization and Adaptation