Instance Adaptive Self-Training for Unsupervised Domain Adaptation
This work addresses the problem of domain shift in deep learning for semantic segmentation, offering a scalable and efficient solution that is incremental in improving pseudo-label quality.
The paper tackles the challenge of unsupervised domain adaptation for semantic segmentation by proposing an instance adaptive self-training framework, which achieves superior performance on benchmarks like GTA5 to Cityscapes and SYNTHIA to Cityscapes compared to state-of-the-art methods.
The divergence between labeled training data and unlabeled testing data is a significant challenge for recent deep learning models. Unsupervised domain adaptation (UDA) attempts to solve such a problem. Recent works show that self-training is a powerful approach to UDA. However, existing methods have difficulty in balancing scalability and performance. In this paper, we propose an instance adaptive self-training framework for UDA on the task of semantic segmentation. To effectively improve the quality of pseudo-labels, we develop a novel pseudo-label generation strategy with an instance adaptive selector. Besides, we propose the region-guided regularization to smooth the pseudo-label region and sharpen the non-pseudo-label region. Our method is so concise and efficient that it is easy to be generalized to other unsupervised domain adaptation methods. Experiments on 'GTA5 to Cityscapes' and 'SYNTHIA to Cityscapes' demonstrate the superior performance of our approach compared with the state-of-the-art methods.