Objective
This article provides guidance on using machine learning-powered transcription tools on YouTube, Microsoft Teams, and Zoom, aiming to enhance video accessibility and meet regulatory requirements.
Audience
Environment
YouTube
Procedure
Those who use Google's YouTube can turn on a closed captioning tool that will use machine learning to add timed text to their respective videos when the 'cc' button is turned on by viewers. Others use this capability on the back end of their videos and then download the .srt files to use in other video hosting platforms.
The U.S. Federal Communications Commission requires a 99% accuracy rate for the wording, spelling punctuation, and grammar, in the transcripts. The voice-to-text transcripts from the uploaded videos may be edited (to correct for mistakes in the auto-transcription) and enriched (with more descriptive information about actions and settings and other relevant non-spoken data).
Note: In Live Microsoft Teams (Office 365) meetings and live Zoom sessions, there is real-time voice-to-text transcription available. Further, both tools enable access to downloadable transcripts post-event, if the proper settings are accessed.
For more information, please see the resources below.
References
Add your own closed captions. (2021). https://support.google.com/youtube/answer/2734796?hl=en
Use automatic captioning. 2021). https://support.google.com/youtube/answer/6373554?hl=en