Adversarial Source Generation for Source-Free Domain Adaptation

Chaoran Cui,Fan’an Meng,Chunyun Zhang,Ziyi Liu,Lei Zhu,Shuai Gong,Xue Lin

DOI: https://doi.org/10.1109/tcsvt.2023.3337796

IF: 5.859

2023-01-01

IEEE Transactions on Circuits and Systems for Video Technology

Abstract:Unsupervised domain adaptation aims to transfer the knowledge learned from a labeled source domain to an unlabeled target domain with different data distributions. However, in practice, source samples are not always available due to privacy protection and storage resource limitations. To address this concern, Source-Free Domain Adaptation (SFDA) has recently attracted growing research attention, as it only needs a pre-trained source model without direct access to source data. In this paper, we propose a novel Adversarial SOurce GEneration (ASOGE) method for SFDA, which introduces an additional generative module to produce synthetic labeled source samples and uses them to facilitate cross-domain adaptation. Unlike early studies that train the generator independently and perform the adaptation only after the generator is finished, ASOGE integrates the generation and adaptation stages within a collaborative framework by making them play an adversarial game. In the generation stage, the labeled source samples are not produced blindly; instead, they are hard-to-align samples that provide knowledge more worth learning for the adaptation stage. To achieve a fine-grained domain alignment, a class-aware discrepancy between source and target domains is measured via contrastive learning. Extensive experiments on benchmark datasets demonstrate the effectiveness of ASOGE compared to the state-of-the-art methods.

engineering, electrical & electronic