Extending Lecture Recording Systems

Adam Reed, Australian National UniversityPart of XW12

I’ve recently done a proof of concept with our lecture recording system to automatically add chapter markers within the video. Once I determined where the chapters were I’ve extracted still frames and run them through OCR. Finally I created a HTML 5 player that would sync the text with the slides. It’s been done using free or open source software. This starts to make video “searchable”, instead of being a large binary blob.

This presentation will outline what I’ve done, the tools I used and how I have put them together. I would then like to open the session up to learn from other attendees what they are doing to add value to video assets, particularly any work around making video accessible, and / or people looking at automatic transcription of audio.

  839 KB