Compact Network Training for Person ReID

TL;DR

This paper introduces a compact, randomly initialized neural network for person re-identification that outperforms state-of-the-art models on key benchmarks and demonstrates versatility in multi-object tracking tasks.

Contribution

A novel small-sized, randomly initialized model and training regime for person ReID, reducing complexity and enabling easier architecture modifications.

Findings

Outperforms SotA on Market1501 and DukeMTMC datasets

Demonstrates robustness and versatility in multi-object tracking

Shows effectiveness with a compact, randomly initialized network

Abstract

The task of person re-identification (ReID) has attracted growing attention in recent years leading to improved performance, albeit with little focus on real-world applications. Most SotA methods are based on heavy pre-trained models, e.g. ResNet50 (~25M parameters), which makes them less practical and more tedious to explore architecture modifications. In this study, we focus on a small-sized randomly initialized model that enables us to easily introduce architecture and training modifications suitable for person ReID. The outcomes of our study are a compact network and a fitting training regime. We show the robustness of the network by outperforming the SotA on both Market1501 and DukeMTMC. Furthermore, we show the representation power of our ReID network via SotA results on a different task of multi-object tracking.