Advertisement
Please join us at GTLUG's monthly meeting--second Monday of the month. As a reminder, we are now meeting at Northwestern Michigan College (NMC), in the lower level of the Timothy J. Nelson Innovation Center (TJNIC) building, Room #15.This month we'll look at Whisper--an machine learning audio transcription tool. It provides speech-to-text capability. We'll discuss how it works and how to compile (for any hardware optimization) and run it on your own local hardware.
While the tool we'll be discussing has built-in subtitle file generation. For long media streams though (i.e. podcasts, TV, movies, old VHS, etc.), Whisper can and will hallucinate. So to assist in generating subtitles for long videos, we'll also cover a separate open project, that our presenter has developed. This other tool will do statistical analysis of the text to find and improve problematic transcriptions.
In theory Whisper can also be used for spoken language translation (specifically to English). One of our members suggested that it might be useful for translating Japanese-dubbed (no subs) anime.
We'll also save a little time of the meeting to touch base on the continuing rocketry project.
If you have any questions, or comments, please post on the event wall, or /join us on IRC (see event host for details).
After the meeting, given sufficient interest (please speak up if you are), we may retire to some establishment nearby for beverages and food.
Advertisement
Event Venue & Nearby Stays
Northwestern Michigan College, Timothy J. Nelson Innovation Center, Room 15, Traverse City, United States
