Class Relationship Embedded Learning for Source-Free Unsupervised Domain Adaptation

• Published in: 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) pp. 7619 - 7629
• Main Authors: Zhang, Yixin, Wang, Zilei, He, Weinan
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.06.2023
• Subjects: Adaptation models; Computational modeling; Computer vision; continual; Data models; low-shot; meta; or long-tail learning; Pattern recognition; Prototypes; Task analysis; Transfer
• ISSN: 2575-7075
• DOI: 10.1109/CVPR52729.2023.00736

This work focuses on a practical knowledge transfer task defined as Source-Free Unsupervised Domain Adaptation (SFUDA), where only a well-trained source model and unlabeled target data are available. To fully utilize source knowledge, we propose to transfer the class relationship, which is domain-invariant but still under-explored in previous works. To this end, we first regard the classifier weights of the source model as class prototypes to compute class relationship, and then propose a novel probability-based similarity between target-domain samples by embedding the source-domain class relationship, resulting in Class Relationship embedded Similarity (CRS). Here the inter-class term is particularly considered in order to more accurately represent the similarity between two samples, in which the source prior of class relationship is utilized by weighting. Finally, we propose to embed CRS into contrastive learning in a unified form. Here both class-aware and instance discrimination contrastive losses are employed, which are complementary to each other. We combine the proposed method with existing representative methods to evaluate its efficacy in multiple SFUDA settings. Extensive experimental results reveal that our method can achieve state-of-the-art performance due to the transfer of domain-invariant class relationship. 1 1 Code is available at https://github.com/zhyx12/CRCo