期刊
IEEE TRANSACTIONS ON IMAGE PROCESSING
卷 31, 期 -, 页码 6951-6963出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TIP.2022.3217697
关键词
Collaboration; Feature extraction; Training; Federated learning; Cameras; Task analysis; Representation learning; Person re-identification; cross-modality; collaborative learning; partially interactive-shared
资金
- National Natural Science Foundation of China [62271484]
- National Science Fund for Distinguished Young Scholars [61925112]
- Innovation Capability Support Program of Shaanxi [2020TD-015]
Visible-infrared person re-identification is a challenging task due to the significant differences between images captured in different spectra. This paper proposes a partially interactive collaboration method to reduce the modality gap, achieving impressive results through the collaborative shallow layers and shared deep layers architecture.
Visible-infrared person re-identification (VI-ReID) task aims to retrieve the same person between visible and infrared images. VI-ReID is challenging as the images captured by different spectra present large cross-modality discrepancy. Many methods adopt a two-stream network and design additional constraint conditions to extract shared features for different modalities. However, the interaction between the feature extraction processes of different modalities is rarely considered. In this paper, a partially interactive collaboration method is proposed to exploit the complementary information of different modalities to reduce the modality gap for VI-ReID. Specifically, the proposed method is achieved in a partially interactive-shared architecture: collaborative shallow layers and shared deep layers. The collaborative shallow layers consider the interaction between modality-specific features of different modalities, encouraging the feature extraction processes of different modalities constrain each other to enhance feature representations. The shared deep layers further embed the modality-specific features to a common space to endow them the same identity discriminability. To ensure the interactive collaborative learning implement effectively, the conventional loss and collaborative loss are utilized jointly to train the whole network. Extensive experiments on two publicly available VI-ReID datasets verify the superiority of the proposed PIC method. Specifically, the proposed method achieves a rank-1 accuracy of 83.6% and 57.5% on RegDB and SYSU-MM01 datasets, respectively.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据