1 citations · 1 across the 1 of their papers we have counts for
1 paper · 1 filter
Jaisidh Singh, Ishaan Shrivastava, Mayank Vatsa +2
Existing vision-language models (VLMs) treat text descriptions as a unit, confusing individual concepts in a prompt and impairing visual semantic matching and reasoning. An importa…