Energy Clustering for Unsupervised Person Re-identification

TL;DR

This paper introduces an energy distance-based clustering method for unsupervised person re-identification, improving clustering quality by considering both inter- and intra-cluster distances, leading to better re-ID performance.

Contribution

It proposes a novel energy distance measure with regularization for hierarchical clustering in unsupervised person re-ID, outperforming existing methods.

Findings

Significant performance improvements on Market-1501, DukeMTMC-reID, and MARS datasets.

Outperforms some transfer learning methods in unsupervised re-ID.

Effective in balancing diversity and similarity in clustering.

Abstract

Due to the high cost of data annotation in supervised learning for person re-identification (Re-ID) methods, unsupervised learning becomes more attractive in the real world. The Bottom-up Clustering (BUC) approach based on hierarchical clustering serves as one promising unsupervised clustering method. One key factor of BUC is the distance measurement strategy. Ideally, the distance measurement should consider both inter-cluster and intra-cluster distance of all samples. However, BUC uses the minimum distance, only considers a pair of the nearest sample between two clusters and ignores the diversity of other samples in clusters. To solve this problem, we propose to use the energy distance to evaluate both the inter-cluster and intra-cluster distance in hierarchical clustering(E-cluster), and use the sum of squares of deviations(SSD) as a regularization term to further balance the diversity and similarity of energy distance evaluation. We evaluate our method on large scale re-ID datasets, including Market-1501, DukeMTMC-reID and MARS. Extensive experiments show that our method obtains significant improvements over the state-of-the-art unsupervised methods, and even better than some transfer learning methods.