1 paper
Dongyun Lin, Yi Cheng, Shangbo Mao +2
Large vision-language models have impressively promote the performance of 2D visual recognition under zero/few-shot scenarios. In this paper, we focus on exploiting the large visio…