A question about this work

Hi, this is an excelent work and the results are impressive. However, there is one question that I'm quite stuck in your paper:

If I understand correctly, keyframes are extracted from the input video. But I wonder how the keyframes are obtained? This seems not to be introduced in the paper.

Sorry to bother you. Hope for your answer.

Start by checking the paper and repository documentation for the keyframe-extraction method used for input video. Confirm where that process is described or implemented, then clarify the explanation so a reader can identify how keyframes are obtained; the issue names no file or test.

Written by the indexing model from the issue text.