Divide and Contrast: Source-free Domain Adaptation via Adaptive Contrastive Learning.
Ziyi ZhangWeikai ChenHui ChengZhen LiSiyuan LiLiang LinGuanbin LiPublished in: NeurIPS (2022)
Keyphrases
- domain adaptation
- multiple sources
- learning process
- learning algorithm
- labeled data
- machine learning
- cross domain
- unsupervised learning
- transfer learning
- transfer knowledge
- learning tasks
- test data
- document classification
- unlabeled data
- supervised learning
- target domain
- prior knowledge
- reinforcement learning
- manifold alignment