One-shot semantic segmentation is to segment the object regions of unseen categories with only one annotated example as the supervision. Existing methods often adopt the multimodal pre-trained model ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results