ToVE: Efficient Vision-Language Learning via Knowledge Transfer from Vision Experts

CLIP
VLM
Re-assign community