Cross-Dataset Person Re-Identification via Unsupervised Pose Disentanglement and Adaptation

TL;DR

This paper introduces PDA-Net, a novel unsupervised method for cross-dataset person re-identification that disentangles pose and domain features to improve generalization across different camera datasets.

Contribution

The paper proposes PDA-Net, an unsupervised pose disentanglement and adaptation network that enhances cross-dataset person re-ID without requiring labeled target data.

Findings

Outperforms state-of-the-art cross-dataset re-ID methods

Effectively disentangles pose and domain features without supervision

Achieves superior results on benchmark datasets

Abstract

Person re-identification (re-ID) aims at recognizing the same person from images taken across different cameras. To address this challenging task, existing re-ID models typically rely on a large amount of labeled training data, which is not practical for real-world applications. To alleviate this limitation, researchers now targets at cross-dataset re-ID which focuses on generalizing the discriminative ability to the unlabeled target domain when given a labeled source domain dataset. To achieve this goal, our proposed Pose Disentanglement and Adaptation Network (PDA-Net) aims at learning deep image representation with pose and domain information properly disentangled. With the learned cross-domain pose invariant feature space, our proposed PDA-Net is able to perform pose disentanglement across domains without supervision in identities, and the resulting features can be applied to cross-dataset re-ID. Both of our qualitative and quantitative results on two benchmark datasets confirm the effectiveness of our approach and its superiority over the state-of-the-art cross-dataset Re-ID approaches.