SWAT: Sliding Window Adversarial Training for Gradual Domain Adaptation
This addresses domain adaptation challenges for machine learning models when faced with gradual but significant domain shifts, representing an incremental advancement in the field.
The paper tackles the problem of steep domain shifts in machine learning by proposing Sliding Window Adversarial Training (SWAT) for Gradual Domain Adaptation, achieving improvements such as 6.1% on Rotated MNIST and 4.1% on CIFAR-100C over previous methods.
Domain shifts are critical issues that harm the performance of machine learning. Unsupervised Domain Adaptation (UDA) mitigates this issue but suffers when the domain shifts are steep and drastic. Gradual Domain Adaptation (GDA) alleviates this problem in a mild way by gradually adapting from the source to the target domain using multiple intermediate domains. In this paper, we propose Sliding Window Adversarial Training (SWAT) for GDA. SWAT first formulates adversarial streams to connect the feature spaces of the source and target domains. Then, a sliding window paradigm is designed that moves along the adversarial stream to gradually narrow the small gap between adjacent intermediate domains. When the window moves to the end of the stream, i.e., the target domain, the domain shift is explicitly reduced. Extensive experiments on six GDA benchmarks demonstrate the significant effectiveness of SWAT, especially 6.1% improvement on Rotated MNIST and 4.1% advantage on CIFAR-100C over the previous methods.