1 paper · 1 filter
Arka Mukherjee, Kaleen Shrestha, Larissa Zhu +1
Modern vision-language models (VLMs) have shown promising results in long-video understanding due to the rich semantic information they can capture. However, most methods focus on…