eX-ViT: A Novel explainable vision transformer for weakly supervised semantic segmentation.
Lu YuWei XiangJuan FangYi-Ping Phoebe ChenLianhua ChiPublished in: Pattern Recognit. (2023)
Keyphrases
- weakly supervised
- semantic segmentation
- superpixels
- object class
- topic models
- computer vision
- relation extraction
- semi supervised
- named entities
- object segmentation
- long range
- scene classification
- multiscale
- bounding box
- segmentation method
- conditional random fields
- segmentation algorithm
- object categories
- natural images
- input image
- pairwise
- image processing