Moving Towards Centers: Re-ranking with Attention and Memory for Re-identification
This work addresses the challenge of post-processing optimization in re-identification for computer vision applications, offering a novel method that significantly outperforms existing re-ranking approaches.
The paper tackles the problem of improving retrieval performance in person and vehicle re-identification by proposing a re-ranking network that predicts correlations between probe and neighbor samples, moving embeddings toward identity centers to enhance cluster compactness. It achieves significant improvements, such as an average 4.83% CMC@1 and 14.83% mAP boost on large-scale datasets like VERI-Wild, MSMT17, and VehicleID.
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.