Open Compound Domain Adaptation with Object Style Compensation for Semantic Segmentation.
Tingliang FengHao ShiXueyang LiuWei FengLiang WanYanlin ZhouDi LinPublished in: CoRR (2023)
Keyphrases
- domain adaptation
- semantic segmentation
- object classes
- superpixels
- cross domain
- labeled data
- d objects
- semi supervised
- target domain
- conditional random fields
- object segmentation
- weakly supervised
- transfer learning
- multiple sources
- object categories
- object class
- bounding box
- unlabeled data
- semi supervised learning
- training data
- data sets
- supervised learning
- pairwise
- image segmentation