Segmentation only uses sparse annotations: Unified weakly and semi-supervised learning in medical images.
Feng GaoMinhao HuMin-Er ZhongShixiang FengXuwei TianXiaochun MengMa-yi-di-li Ni-jia-tiZeping HuangMinyi LvTao SongXiaofan ZhangXiaoguang ZouXiaojian WuPublished in: Medical Image Anal. (2022)
Keyphrases
- medical images
- semi supervised learning
- semi supervised
- unlabeled data
- medical imaging
- anatomical structures
- labeled data
- medical image segmentation
- deformable models
- supervised learning
- medical image analysis
- magnetic resonance images
- segmentation of medical images
- unsupervised learning
- partially labeled
- mr images
- semi supervised classification
- training data
- co training
- biomedical images
- active learning
- machine learning
- semantic annotation
- high dimensional
- data sets
- sparse representation
- shape prior
- multi view
- random projections
- medical image retrieval
- computer vision