Instance Adaptive Self-Training for Unsupervised Domain Adaptation

Abstract

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.

ECCV 2020

References in corpus (5)

• ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation
• Conditional Adversarial Domain Adaptation
• Regularization for Deep Learning: A Taxonomy
• Category Anchor-Guided Unsupervised Domain Adaptation for Semantic Segmentation
• Constructing Self-motivated Pyramid Curriculums for Cross-Domain Semantic Segmentation: A Non-Adversarial Approach

Cited by in corpus (4)

• Learning Invariant Representations and Risks for Semi-supervised Domain Adaptation
• Unsupervised Domain Adaptation for Video Semantic Segmentation
• LabOR: Labeling Only if Required for Domain Adaptive Semantic Segmentation
• Contrastive Learning and Self-Training for Unsupervised Domain Adaptation in Semantic Segmentation