Abstract: We present a fully automatic system for extracting the semantic structure of a typical academic presentation video, which captures the whole presentation stage with abundant camera motions ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback