ResCLIP: Residual Attention for Training-free Dense Vision-language Inference

Devs

ResCLIP: Residual Attention for Training-free Dense Vision-language Inference | Read Paper on Bytez