When watching tutorial videos, I always wanted to just drag a part of the video, ask a question about it, and get an answer. In our upcoming
#CHI2024 paper, we studied users' question asking behavior and designed a pipeline that generates answers to questions with visual anchors.