Unsupervised Person Re-identification via Multi-label Classification

TL;DR

This paper introduces an unsupervised person re-identification method that formulates the task as multi-label classification, using iterative label prediction and a memory-based loss to improve discriminative feature learning without true labels.

Contribution

It proposes a novel multi-label classification framework with a memory-based loss and cycle consistency for unsupervised person ReID, achieving state-of-the-art results and enabling transfer learning.

Findings

Outperforms existing unsupervised ReID methods on large-scale datasets.

Effectively leverages unlabeled data for discriminative feature learning.

Achieves state-of-the-art performance in transfer learning scenarios.

Abstract

The challenge of unsupervised person re-identification (ReID) lies in learning discriminative features without true labels. This paper formulates unsupervised person ReID as a multi-label classification task to progressively seek true labels. Our method starts by assigning each person image with a single-class label, then evolves to multi-label classification by leveraging the updated ReID model for label prediction. The label prediction comprises similarity computation and cycle consistency to ensure the quality of predicted labels. To boost the ReID model training efficiency in multi-label classification, we further propose the memory-based multi-label classification loss (MMCL). MMCL works with memory-based non-parametric classifier and integrates multi-label classification and single-label classification in a unified framework. Our label prediction and MMCL work iteratively and substantially boost the ReID performance. Experiments on several large-scale person ReID datasets demonstrate the superiority of our method in unsupervised person ReID. Our method also allows to use labeled person images in other domains. Under this transfer learning setting, our method also achieves state-of-the-art performance.