Computationally-Efficient Vision Transformer for Medical Image Semantic Segmentation Via Dual Pseudo-Label Supervision.
Ziyang WangNanqing DongIrina VoiculescuPublished in: ICIP (2022)
Keyphrases
- medical images
- semantic segmentation
- superpixels
- conditional random fields
- computer vision
- weakly supervised
- scene classification
- deformable models
- pascal voc
- object classes
- object categories
- active learning
- object class
- region of interest
- image understanding
- object recognition
- image processing
- multi label
- bounding box
- image set
- class labels
- image classification
- co occurrence
- multiscale