Visual Transcripts: Lecture Notes from Blackboard-Style Lecture Videos

摘要

Blackboard-style lecture videos are popular, but learning using existingrnvideo player interfaces can be challenging. Viewers cannotrnconsume the lecture material at their own pace, and the content isrnalso difficult to search or skim. For these reasons, some people preferrnlecture notes to videos. To address these limitations, we presentrnVisual Transcripts, a readable representation of lecture videos thatrncombines visual information with transcript text. To generate a VisualrnTranscript, we first segment the visual content of a lecture intorndiscrete visual entities that correspond to equations, figures, or linesrnof text. Then, we analyze the temporal correspondence between therntranscript and visuals to determine how sentences relate to visualrnentities. Finally, we arrange the text and visuals in a linear layoutrnbased on these relationships. We compare our result with a standardrnvideo player, and a state-of-the-art interface designed specificallyrnfor blackboard-style lecture videos. User evaluation suggestsrnthat users prefer our interface for learning and that our interface isrneffective in helping them browse or search through lecture videos.