Target-Relevant Knowledge Preservation for Multi-Source Domain Adaptive Object Detection.
Jiaxi WuJiaxin ChenMengzhe HeYiru WangBo LiBingqi MaWeihao GanWei WuYali WangDi HuangPublished in: CoRR (2022)
Keyphrases
- relevant knowledge
- object detection
- target domain
- source domain
- domain adaptation
- cross domain
- transfer learning
- labeled data
- data distribution
- multiple sources
- knowledge base
- test data
- training samples
- computer vision
- concept space
- sentiment classification
- text categorization
- data sets
- unlabeled data
- reinforcement learning
- training data
- decision trees
- databases