Fetching the paper…

TagCLIP: Improving Discrimination Ability of Open-Vocabulary Semantic Segmentation · Around