Margin Sample Mining Loss: A Deep Learning Based Method for Person Re-identification

TL;DR

This paper introduces Margin Sample Mining Loss (MSML), a novel deep metric learning loss function that improves person re-identification accuracy by effectively mining hard samples, outperforming existing methods on multiple datasets.

Contribution

The paper proposes MSML, a new metric learning loss with hard sample mining, demonstrating superior performance over existing losses like triplet loss in person ReID tasks.

Findings

MSML outperforms state-of-the-art algorithms on Market1501, MARS, CUHK03, and CUHK-SYSU datasets.

MSML achieves higher accuracy in person re-identification tasks.

Experimental results validate the effectiveness of MSML in deep learning frameworks.

Abstract

Person re-identification (ReID) is an important task in computer vision. Recently, deep learning with a metric learning loss has become a common framework for ReID. In this paper, we also propose a new metric learning loss with hard sample mining called margin smaple mining loss (MSML) which can achieve better accuracy compared with other metric learning losses, such as triplet loss. In experi- ments, our proposed methods outperforms most of the state-of-the-art algorithms on Market1501, MARS, CUHK03 and CUHK-SYSU.