Video Temporal Relationship Mining for Data-Efficient Person Re-identification

TL;DR

This paper introduces a novel post-processing method for person re-identification that leverages temporal relationships in video sequences to improve retrieval accuracy by iteratively refining matches.

Contribution

It proposes a new video temporal relationship mining strategy that enhances re-identification accuracy through iterative image retrieval based on temporal continuity.

Findings

Improved retrieval accuracy in person re-identification tasks.

Effective use of temporal relationships in video sequences.

Robust retrieval sequences achieved through iterative search.

Abstract

This paper is a technical report to our submission to the ICCV 2021 VIPriors Re-identification Challenge. In order to make full use of the visual inductive priors of the data, we treat the query and gallery images of the same identity as continuous frames in a video sequence. And we propose one novel post-processing strategy for video temporal relationship mining, which not only calculates the distance matrix between query and gallery images, but also the matrix between gallery images. The initial query image is used to retrieve the most similar image from the gallery, then the retrieved image is treated as a new query to retrieve its most similar image from the gallery. By iteratively searching for the closest image, we can achieve accurate image retrieval and finally obtain a robust retrieval sequence.