Towards Stable and Comprehensive Domain Alignment: Max-Margin Domain-Adversarial Training

Open in new window