MetaSeg: Content-Aware Meta-Net for Omni-Supervised Semantic Segmentation.
Shenwang JiangJianan LiYing WangWenxuan WuJizhou ZhangBo HuangTingfa XuPublished in: CoRR (2024)
Keyphrases
- content aware
- semantic segmentation
- label transfer
- weakly supervised
- superpixels
- conditional random fields
- content delivery
- scene classification
- object categories
- pascal voc
- object classes
- semi supervised
- object detection
- learning algorithm
- topic models
- object class
- multiscale
- video sequences
- machine learning
- image analysis
- object recognition
- feature selection