🤖 AI Summary
To address insufficient fine-grained discriminative feature modeling in person re-identification (ReID), this paper proposes LAGA-Net, an end-to-end learnable multi-branch attention network. Methodologically, it introduces a novel local-aware global attention mechanism that jointly integrates spatial, channel-wise, global, and local features; incorporates relative positional encoding to enhance the robustness of spatial attention to pixel layout variations; and designs an unsupervised horizontal soft stripe segmentation module, eliminating reliance on external supervision such as pose estimation. LAGA-Net achieves state-of-the-art performance on four mainstream person ReID benchmarks (Market-1501, DukeMTMC-reID, CUHK03-NP, MSMT17) and two hand ReID datasets. Ablation studies confirm the substantial contributions of each component, with the local-aware attention mechanism and relative positional encoding proving particularly critical for performance gains.
📝 Abstract
Learning representative, robust and discriminative information from images is essential for effective person re-identification (Re-Id). In this paper, we propose a compound approach for end-to-end discriminative deep feature learning for person Re-Id based on both body and hand images. We carefully design the Local-Aware Global Attention Network (LAGA-Net), a multi-branch deep network architecture consisting of one branch for spatial attention, one branch for channel attention, one branch for global feature representations and another branch for local feature representations. The attention branches focus on the relevant features of the image while suppressing the irrelevant backgrounds. In order to overcome the weakness of the attention mechanisms, equivariant to pixel shuffling, we integrate relative positional encodings into the spatial attention module to capture the spatial positions of pixels. The global branch intends to preserve the global context or structural information. For the the local branch, which intends to capture the fine-grained information, we perform uniform partitioning to generate stripes on the conv-layer horizontally. We retrieve the parts by conducting a soft partition without explicitly partitioning the images or requiring external cues such as pose estimation. A set of ablation study shows that each component contributes to the increased performance of the LAGA-Net. Extensive evaluations on four popular body-based person Re-Id benchmarks and two publicly available hand datasets demonstrate that our proposed method consistently outperforms existing state-of-the-art methods.
