Attention-based Few-Shot Person Re-identification Using Meta Learning

TL;DR

This paper introduces ARM, an attention-based meta-learning model for few-shot person re-identification, effectively handling new identities with limited examples by leveraging a novel feature encoding scheme and meta-learning framework.

Contribution

The paper proposes a new attention-based feature encoding scheme within a meta-learning framework specifically designed for few-shot person re-identification, improving discriminative power and adaptability.

Findings

ARM outperforms state-of-the-art on multiple datasets

Effective handling of new identities with few examples

Enhanced discriminative feature encoding scheme

Abstract

In this paper, we investigate the challenging task of person re-identification from a new perspective and propose an end-to-end attention-based architecture for few-shot re-identification through meta-learning. The motivation for this task lies in the fact that humans, can usually identify another person after just seeing that given person a few times (or even once) by attending to their memory. On the other hand, the unique nature of the person re-identification problem, i.e., only few examples exist per identity and new identities always appearing during testing, calls for a few shot learning architecture with the capacity of handling new identities. Hence, we frame the problem within a meta-learning setting, where a neural network based meta-learner is trained to optimize a learner i.e., an attention-based matching function. Another challenge of the person re-identification problem is the small inter-class difference between different identities and large intra-class difference of the same identity. In order to increase the discriminative power of the model, we propose a new attention-based feature encoding scheme that takes into account the critical intra-view and cross-view relationship of images. We refer to the proposed Attention-based Re-identification Metalearning model as ARM. Extensive evaluations demonstrate the advantages of the ARM as compared to the state-of-the-art on the challenging PRID2011, CUHK01, CUHK03 and Market1501 datasets.