PAC-GAN: An Effective Pose Augmentation Scheme for Unsupervised Cross-View Person Re-identification

TL;DR

This paper introduces PAC-GAN, an unsupervised pose augmentation framework using conditional GANs to generate pose-rich samples, significantly improving cross-view person re-identification performance by addressing dataset limitations.

Contribution

The paper proposes a novel unsupervised pose augmentation scheme with a CGAN-based framework, enhancing data diversity for better re-identification accuracy.

Findings

PAC-GAN outperforms existing methods on benchmark datasets.

Pose augmentation improves recognition accuracy in unsupervised settings.

The approach effectively increases dataset diversity with synthesized pose-rich samples.

Abstract

Person re-identification (person Re-Id) aims to retrieve the pedestrian images of a same person that captured by disjoint and non-overlapping cameras. Lots of researchers recently focuse on this hot issue and propose deep learning based methods to enhance the recognition rate in a supervised or unsupervised manner. However, two limitations that cannot be ignored: firstly, compared with other image retrieval benchmarks, the size of existing person Re-Id datasets are far from meeting the requirement, which cannot provide sufficient pedestrian samples for the training of deep model; secondly, the samples in existing datasets do not have sufficient human motions or postures coverage to provide more priori knowledges for learning. In this paper, we introduce a novel unsupervised pose augmentation cross-view person Re-Id scheme called PAC-GAN to overcome these limitations. We firstly present the formal definition of cross-view pose augmentation and then propose the framework of PAC-GAN that is a novel conditional generative adversarial network (CGAN) based approach to improve the performance of unsupervised corss-view person Re-Id. Specifically, The pose generation model in PAC-GAN called CPG-Net is to generate enough quantity of pose-rich samples from original image and skeleton samples. The pose augmentation dataset is produced by combining the synthesized pose-rich samples with the original samples, which is fed into the corss-view person Re-Id model named Cross-GAN. Besides, we use weight-sharing strategy in the CPG-Net to improve the quality of new generated samples. To the best of our knowledge, we are the first try to enhance the unsupervised cross-view person Re-Id by pose augmentation, and the results of extensive experiments show that the proposed scheme can combat the state-of-the-arts.