Moving Towards Centers: Re-ranking with Attention and Memory for Re-identification

TL;DR

This paper introduces a novel re-ranking network for person and vehicle re-identification that leverages attention and memory mechanisms to predict correlations and improve retrieval accuracy, significantly surpassing existing methods.

Contribution

It proposes a re-ranking approach using Transformer-based contextual aggregation and memory modules to enhance feature embeddings and correlation prediction in re-ID tasks.

Findings

Outperforms state-of-the-art re-ranking methods on multiple benchmarks.

Achieves an average 4.83% CMC@1 and 14.83% mAP improvements.

Effective in large-scale person and vehicle re-ID datasets.

Abstract

Re-ranking utilizes contextual information to optimize the initial ranking list of person or vehicle re-identification (re-ID), which boosts the retrieval performance at post-processing steps. This paper proposes a re-ranking network to predict the correlations between the probe and top-ranked neighbor samples. Specifically, all the feature embeddings of query and gallery images are expanded and enhanced by a linear combination of their neighbors, with the correlation prediction serving as discriminative combination weights. The combination process is equivalent to moving independent embeddings toward the identity centers, improving cluster compactness. For correlation prediction, we first aggregate the contextual information for probe's k-nearest neighbors via the Transformer encoder. Then, we distill and refine the probe-related features into the Contextual Memory cell via attention mechanism. Like humans that retrieve images by not only considering probe images but also memorizing the retrieved ones, the Contextual Memory produces multi-view descriptions for each instance. Finally, the neighbors are reconstructed with features fetched from the Contextual Memory, and a binary classifier predicts their correlations with the probe. Experiments on six widely-used person and vehicle re-ID benchmarks demonstrate the effectiveness of the proposed method. Especially, our method surpasses the state-of-the-art re-ranking approaches on large-scale datasets by a significant margin, i.e., with an average 4.83% CMC@1 and 14.83% mAP improvements on VERI-Wild, MSMT17, and VehicleID datasets.