Domain Adaptation for Semantic Segmentation via Patch-Wise Contrastive Learning
This addresses the problem of domain shift in semantic segmentation for computer vision applications, offering a more efficient training method with significant annotation cost reductions.
The paper tackles unsupervised and semi-supervised domain adaptation for semantic segmentation by using patch-wise contrastive learning to align features across domains, resulting in improved performance over state-of-the-art methods, especially with few target annotations, and achieving up to 75% annotation cost savings in weakly-supervised settings.
We introduce a novel approach to unsupervised and semi-supervised domain adaptation for semantic segmentation. Unlike many earlier methods that rely on adversarial learning for feature alignment, we leverage contrastive learning to bridge the domain gap by aligning the features of structurally similar label patches across domains. As a result, the networks are easier to train and deliver better performance. Our approach consistently outperforms state-of-the-art unsupervised and semi-supervised methods on two challenging domain adaptive segmentation tasks, particularly with a small number of target domain annotations. It can also be naturally extended to weakly-supervised domain adaptation, where only a minor drop in accuracy can save up to 75% of annotation cost.