MSINet: Twins Contrastive Search of Multi-Scale Interaction for Object ReID
This work addresses the challenge of improving retrieval performance in object ReID, which is important for surveillance and security applications, by introducing a novel NAS approach tailored to ReID-specific training schemes.
The authors tackled the problem of object re-identification (ReID) by proposing a Neural Architecture Search (NAS) scheme that includes a Twins Contrastive Mechanism to better simulate real-world training and a Multi-Scale Interaction search space, resulting in MSINet, which surpasses state-of-the-art methods in both in-domain and cross-domain scenarios.
Neural Architecture Search (NAS) has been increasingly appealing to the society of object Re-Identification (ReID), for that task-specific architectures significantly improve the retrieval performance. Previous works explore new optimizing targets and search spaces for NAS ReID, yet they neglect the difference of training schemes between image classification and ReID. In this work, we propose a novel Twins Contrastive Mechanism (TCM) to provide more appropriate supervision for ReID architecture search. TCM reduces the category overlaps between the training and validation data, and assists NAS in simulating real-world ReID training schemes. We then design a Multi-Scale Interaction (MSI) search space to search for rational interaction operations between multi-scale features. In addition, we introduce a Spatial Alignment Module (SAM) to further enhance the attention consistency confronted with images from different sources. Under the proposed NAS scheme, a specific architecture is automatically searched, named as MSINet. Extensive experiments demonstrate that our method surpasses state-of-the-art ReID methods on both in-domain and cross-domain scenarios. Source code available in https://github.com/vimar-gu/MSINet.