HRDA
Explore how HRDA uses multi-resolution training to improve unsupervised domain adaptation by combining high-resolution for detail and low-resolution for context, enhancing performance on datasets like GTA→Cityscapes. Learn about HRDA's advancements in domain generalization and its impact on state-of-the-art results.
