Dual frame-level and region-level alignment for unsupervised video domain adaptation.
Xinyue HuYingying ZhuPublished in: Neurocomputing (2023)
Keyphrases
- domain adaptation
- manifold alignment
- region level
- semi supervised
- video frames
- pixel level
- object level
- cross domain
- labeled data
- transfer learning
- multiple sources
- video data
- unsupervised learning
- video sequences
- instance level
- semi supervised learning
- training data
- key frames
- target domain
- test data
- supervised learning
- semantic concepts
- unlabeled data
- co training
- sentiment classification
- image segmentation
- dynamic scenes
- learning algorithm
- natural language processing
- pairwise
- decision trees
- computer vision