xiangjieSui/LMM_VQA

Probing the visual quality assessment capability of large multimodal models