Unsupervised Contrastive Domain Adaptation for Semantic Segmentation
This addresses domain adaptation for semantic segmentation, which is crucial for applications like autonomous driving, but it appears incremental as it builds on existing contrastive learning methods.
The paper tackles the problem of semantic segmentation models failing to generalize due to domain shift by introducing contrastive learning for feature alignment and a label expansion approach, achieving 60.2% mIoU on Cityscapes when trained on synthetic GTA5 data with unlabeled Cityscapes images.
Semantic segmentation models struggle to generalize in the presence of domain shift. In this paper, we introduce contrastive learning for feature alignment in cross-domain adaptation. We assemble both in-domain contrastive pairs and cross-domain contrastive pairs to learn discriminative features that align across domains. Based on the resulting well-aligned feature representations we introduce a label expansion approach that is able to discover samples from hard classes during the adaptation process to further boost performance. The proposed approach consistently outperforms state-of-the-art methods for domain adaptation. It achieves 60.2% mIoU on the Cityscapes dataset when training on the synthetic GTA5 dataset together with unlabeled Cityscapes images.