2 citations · 3 across the 13 of their papers we have counts for
1 paper · 2 filters
Zhixiang Wei, Yi Li, Zhehan Kan +38
Despite the significant advancements represented by Vision-Language Models (VLMs), current architectures often exhibit limitations in retaining fine-grained visual information, lea…