A Semi-Supervised Maximum Margin Metric Learning Approach for Small Scale Person Re-identification

TL;DR

This paper introduces a semi-supervised metric learning method for small-scale person re-identification that leverages unlabelled data and addresses the small sample size problem, achieving state-of-the-art results.

Contribution

It proposes a novel semi-supervised approach that combines maximum margin metric learning with high-dimensional kernel mappings and pseudo-class mining for improved re-identification.

Findings

Achieves state-of-the-art performance on four datasets.

Effectively utilizes unlabelled data with limited labeled samples.

Reduces within-class variance to improve discrimination.

Abstract

In video surveillance, person re-identification is the task of searching person images in non-overlapping cameras. Though supervised methods for person re-identification have attained impressive performance, obtaining large scale cross-view labeled training data is very expensive. However, unlabelled data is available in abundance. In this paper, we propose a semi-supervised metric learning approach that can utilize information in unlabelled data with the help of a few labelled training samples. We also address the small sample size problem that inherently occurs due to the few labeled training data. Our method learns a discriminative space where within class samples collapse to singular points, achieving the least within class variance, and then use a maximum margin criterion over a high dimensional kernel space to maximally separate the distinct class samples. A maximum margin criterion with two levels of high dimensional mappings to kernel space is used to obtain better cross-view discrimination of the identities. Cross-view affinity learning with reciprocal nearest neighbor constraints is used to mine new pseudo-classes from the unlabelled data and update the distance metric iteratively. We attain state-of-the-art performance on four challenging datasets with a large margin.