Fetching the paper…

SgVA-CLIP: Semantic-guided Visual Adapting of Vision-Language Models for Few-shot Image Classification · Around