DotSCN: Group Re-identification via Domain-Transferred Single and Couple Representation Learning

TL;DR

This paper introduces TSCN, a deep learning framework for group re-identification that transfers knowledge from labeled person ReID datasets and learns discriminative single and couple representations to handle appearance and layout changes.

Contribution

The paper proposes a novel domain transfer approach and a couple representation learning method to improve group re-identification accuracy significantly.

Findings

Outperforms state-of-the-art by 11.7% on Road Group dataset

Outperforms state-of-the-art by 39.0% on DukeMCMT dataset

Effectively handles appearance and layout variations in G-ReID

Abstract

Group re-identification (G-ReID) is an important yet less-studied task. Its challenges not only lie in appearance changes of individuals which have been well-investigated in general person re-identification (ReID), but also derive from group layout and membership changes. So the key task of G-ReID is to learn representations robust to such changes. To address this issue, we propose a Transferred Single and Couple Representation Learning Network (TSCN). Its merits are two aspects: 1) Due to the lack of labelled training samples, existing G-ReID methods mainly rely on unsatisfactory hand-crafted features. To gain the superiority of deep learning models, we treat a group as multiple persons and transfer the domain of a labeled ReID dataset to a G-ReID target dataset style to learn single representations. 2) Taking into account the neighborhood relationship in a group, we further propose learning a novel couple representation between two group members, that achieves more discriminative power in G-ReID tasks. In addition, an unsupervised weight learning method is exploited to adaptively fuse the results of different views together according to result patterns. Extensive experimental results demonstrate the effectiveness of our approach that significantly outperforms state-of-the-art methods by 11.7\% CMC-1 on the Road Group dataset and by 39.0\% CMC-1 on the DukeMCMT dataset.