1 paper
Jisheng Dang, Huilin Song, Junbin Xiao +6
Grounded Video Question Answering (Grounded VideoQA) requires aligning textual answers with explicit visual evidence. However, modern multimodal models often rely on linguistic pri…