Pluggable Weakly-Supervised Cross-View Learning for Accurate Vehicle Re-Identification

TL;DR

This paper introduces a pluggable weakly-supervised method for vehicle re-identification that learns cross-view consistent features without viewpoint annotations, significantly improving performance across multiple benchmarks.

Contribution

The proposed WCVL module enables cross-view learning without viewpoint labels and can be integrated into existing baselines without re-training.

Findings

Significant performance improvements on four benchmark datasets.

Effective learning of cross-view features using only vehicle IDs.

Seamless integration with existing vehicle ReID models.

Abstract

Learning cross-view consistent feature representation is the key for accurate vehicle Re-identification (ReID), since the visual appearance of vehicles changes significantly under different viewpoints. To this end, most existing approaches resort to the supervised cross-view learning using extensive extra viewpoints annotations, which however, is difficult to deploy in real applications due to the expensive labelling cost and the continous viewpoint variation that makes it hard to define discrete viewpoint labels. In this study, we present a pluggable Weakly-supervised Cross-View Learning (WCVL) module for vehicle ReID. Through hallucinating the cross-view samples as the hardest positive counterparts in feature domain, we can learn the consistent feature representation via minimizing the cross-view feature distance based on vehicle IDs only without using any viewpoint annotation. More importantly, the proposed method can be seamlessly plugged into most existing vehicle ReID baselines for cross-view learning without re-training the baselines. To demonstrate its efficacy, we plug the proposed method into a bunch of off-the-shelf baselines and obtain significant performance improvement on four public benchmark datasets, i.e., VeRi-776, VehicleID, VRIC and VRAI.