Progressive Learning Algorithm for Efficient Person Re-Identification

TL;DR

This paper introduces a Progressive Learning Algorithm (PLA) that enhances person re-identification by combining triplet and cross-entropy losses, achieving high accuracy with reduced model complexity suitable for large-scale applications.

Contribution

The paper proposes a novel progressive learning strategy using Bayesian optimization to improve feature embeddings efficiently for person ReID.

Findings

Achieves 94.7% Rank-1 accuracy on Market-1501

Reduces model parameters by at least 30% compared to state-of-the-art models

Performs comparably or better on three large-scale datasets

Abstract

This paper studies the problem of Person Re-Identification (ReID)for large-scale applications. Recent research efforts have been devoted to building complicated part models, which introduce considerably high computational cost and memory consumption, inhibiting its practicability in large-scale applications. This paper aims to develop a novel learning strategy to find efficient feature embeddings while maintaining the balance of accuracy and model complexity. More specifically, we find by enhancing the classical triplet loss together with cross-entropy loss, our method can explore the hard examples and build a discriminant feature embedding yet compact enough for large-scale applications. Our method is carried out progressively using Bayesian optimization, and we call it the Progressive Learning Algorithm (PLA). Extensive experiments on three large-scale datasets show that our PLA is comparable or better than the-state-of-the-arts. Especially, on the challenging Market-1501 dataset, we achieve Rank-1=94.7\%/mAP=89.4\% while saving at least 30\% parameters than strong part models.