Label Anything: Multi-Class Few-Shot Semantic Segmentation with Visual Prompts
Paper
•
2407.02075
•
Published
Label Anything
Label Anything performs multi-class few-shot semantic segmentation using visual prompts (bboxes, points, masks)