Density-Adaptive Kernel based Efficient Reranking Approaches for Person Reidentification

TL;DR

This paper introduces density-adaptive kernel based reranking methods for person reidentification that improve accuracy while reducing computational costs, by exploiting local density information and incorporating extra probe samples.

Contribution

It proposes novel inverse and bidirectional density-adaptive kernel reranking methods that are efficient, effective, and capable of leveraging additional probe samples for improved performance.

Findings

Effective on six benchmark datasets.

Achieves significant performance improvements.

Reduces computational costs compared to existing methods.

Abstract

Person reidentification (ReID) refers to the task of verifying the identity of a pedestrian observed from nonoverlapping views in a surveillance camera network. It has recently been validated that reranking can achieve remarkable performance improvements in person ReID systems. However, current reranking approaches either require feedback from users or suffer from burdensome computational costs. In this paper, we propose to exploit a density-adaptive smooth kernel technique to achieve efficient and effective reranking. Specifically, we adopt a smooth kernel function to formulate the neighbor relationships among data samples with a density-adaptive parameter. Based on this new formulation, we present two simple yet effective reranking methods, termed \emph{inverse} density-adaptive kernel based reranking (inv-DAKR) and \emph{bidirectional} density-adaptive kernel based reranking (bi-DAKR), in which the local density information in the vicinity of each gallery sample is elegantly exploited. Moreover, we extend the proposed inv-DAKR and bi-DAKR methods to incorporate the available extra probe samples and demonstrate that when and why these extra probe samples are able to improve the local neighborhood and thus further refine the ranking results. Extensive experiments are conducted on six benchmark datasets, including: PRID450s, VIPeR, CUHK03, GRID, Market-1501 and Mars. The experimental results demonstrate that our proposals are effective and efficient.