AttributeNet: Attribute Enhanced Vehicle Re-Identification

TL;DR

AttributeNet enhances vehicle re-identification by jointly extracting identity and attribute features, and introducing a constraint to improve feature discrimination, achieving state-of-the-art results on challenging datasets.

Contribution

The paper proposes AttributeNet, a novel framework that effectively integrates attribute features into vehicle re-identification, with a new Amelioration Constraint to boost discriminative power.

Findings

Achieves state-of-the-art performance on three datasets.

Effectively integrates attribute features with identity features.

The Amelioration Constraint improves feature discrimination.

Abstract

Vehicle Re-Identification (V-ReID) is a critical task that associates the same vehicle across images from different camera viewpoints. Many works explore attribute clues to enhance V-ReID; however, there is usually a lack of effective interaction between the attribute-related modules and final V-ReID objective. In this work, we propose a new method to efficiently explore discriminative information from vehicle attributes (for instance, color and type). We introduce AttributeNet (ANet) that jointly extracts identity-relevant features and attribute features. We enable the interaction by distilling the ReID-helpful attribute feature and adding it into the general ReID feature to increase the discrimination power. Moreover, we propose a constraint, named Amelioration Constraint (AC), which encourages the feature after adding attribute features onto the general ReID feature to be more discriminative than the original general ReID feature. We validate the effectiveness of our framework on three challenging datasets. Experimental results show that our method achieves the state-of-the-art performance.