Towards generalizable person re-identification with a bi-stream generative model. (December 2022)

• Record Type: Journal Article
• Authors: Xu, Xin; Liu, Wei; Wang, Zheng; Hu, Ruimin; Tian, Qi
• Is Part Of: Pattern recognition. Volume 132(2022)
• Publication Date: 2022-12
• Subjects: Person re-identification -- Generalizable re-ID -- Camera-Camera problem -- Camera-Person problem; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2022.108954
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: We decouple the difficulties affecting the person re-identification task into the Camera-Camera (CC) problem and the Camera-Person (CP) problem. We propose a bi-stream generative model for solving the CC and CP problems separately, with promising results. We design a part-weighted loss based on the unbalanced number of human body parts in the dataset to guide the model to focus on the more important parts. Abstract: Generalizable person re-identification (re-ID) has attracted growing attention due to its powerful adaptation capability in the unseen data domain. However, existing solutions often neglect either crossing cameras ( e.g., illumination and resolution differences) or pedestrian misalignments ( e.g., viewpoint and pose discrepancies), which easily leads to poor generalization capability when adapted to the new domain. In this paper, we formulate these difficulties as: 1) Camera-Camera ( CC ) problem, which denotes the various human appearance changes caused by different cameras; 2) Camera-Person ( CP ) problem, which indicates the pedestrian misalignments caused by the same identity person under different camera viewpoints or changing pose. To solve the above issues, we propose a Bi-stream Generative Model (BGM) to learn the fine-grained representations fused with camera-invariant global feature and pedestrian-aligned local feature, which contains an encoding network and two stream decoding sub-network. Guided by original pedestrian images, one stream is employed to learn a camera-invariant global feature for the CC problem via filtering cross-camera interference factors. For the CP problem, another stream learns a pedestrian-aligned local feature for pedestrian alignment using information-complete densely semantically aligned part maps. Moreover, a part-weighted loss function is presented to reduce the influence of missing parts on pedestrian alignment. Extensive experiments demonstrate that our method outperforms the state-of-the-art methods on the large-scale generalizable re-ID benchmarks, involving domain generalization setting and cross-domain setting.