Anti-Forgetting Adaptation for Unsupervised Person Re-identification

TL;DR

This paper introduces a novel framework for unsupervised person re-identification that incrementally adapts to new domains while effectively preventing forgetting of previously learned knowledge, enhancing generalization across multiple domains.

Contribution

The proposed DJAA framework utilizes prototype and instance-level consistency with memory buffers to mitigate forgetting during incremental domain adaptation in person ReID.

Findings

Significantly improves anti-forgetting ability in unsupervised ReID.

Enhances model generalization to unseen domains.

Demonstrates superior performance over existing methods in experiments.

Abstract

Regular unsupervised domain adaptive person re-identification (ReID) focuses on adapting a model from a source domain to a fixed target domain. However, an adapted ReID model can hardly retain previously-acquired knowledge and generalize to unseen data. In this paper, we propose a Dual-level Joint Adaptation and Anti-forgetting (DJAA) framework, which incrementally adapts a model to new domains without forgetting source domain and each adapted target domain. We explore the possibility of using prototype and instance-level consistency to mitigate the forgetting during the adaptation. Specifically, we store a small number of representative image samples and corresponding cluster prototypes in a memory buffer, which is updated at each adaptation step. With the buffered images and prototypes, we regularize the image-to-image similarity and image-to-prototype similarity to rehearse old knowledge. After the multi-step adaptation, the model is tested on all seen domains and several unseen domains to validate the generalization ability of our method. Extensive experiments demonstrate that our proposed method significantly improves the anti-forgetting, generalization and backward-compatible ability of an unsupervised person ReID model.