Gallantries Implements Automated TTS for Slides

Published: Feb 2, 2021

videos  gtn  training  pandemic  remote-teaching  

This past year has been a struggle for all of us, but one task has become apparent as one of the most time consuming portions of course preparation during a pandemic: Video Production.

We have found ourselves spending countless hours recording video lectures, producing videos of us giving slide decks, and videos of us giving trainings on how to run tools in Galaxy. Worse, whenever changes were made we would have to re-record lessons!

Inspired by the efforts of Delphine Lariviere in her video-lectures series, Helena Rasche of the Gallantries has implemented a similar feature of automatic Text-to-Speech (TTS) in the Galaxy Training Network (GTN).

The GTN had existing infrastructure for producing slide decks from easy-to-maintain markdown documents, and, key to the success, existing syntax for writing “Speaker Notes” on slides. These were intended to be little reminders to instructors giving slide decks about what they should say on each slide.

Thanks to this existing groundwork, after we implemented TTS, contributors only had to add video: true to the top of their slides, and they would opt-in to automatic video production. Now our contributors could easily produce slide deck recordings with very little overhead. Best of all, anytime a contributor updated their slide decks, the videos could be re-built as well. As we know the text that would be spoken we could also trivially add accurate subtitles, ensuring accessibility for our deaf/Deaf/HoH community members.

All of this process is completely automated, from building the videos, to uploading them to our S3 bucket and including them in the GTN. Since implementation in November, we have added 16 videos to the GTN.