Transcription (Speech to Text)

[acrawford73]

Transcription APIs are getting much easier to use these days (and affordable). For example, Deepgram API transcribes audio and video files with impressive accuracy. I would like to recommend support for subtitle generation / transcription:

1. Use Deepgram API to auto-generate WebVTT subtitles for all uploaded media, based on Deepgram transcription options set in the django admin.
2. Use Deepgram API to auto-generate WebVTT subtitles for selected uploaded media, a 'Generate Subtitles' button on the EDIT MEDIA page, when global transcription is disabled.
3. This could be further extended to transcribe (speech to text) all or selected video/audio files.
4. Leverage ffmpeg to generate a mono audio file of the video. Perform the transcoding / WebVTT generation by making a Deepgram API request with the URL to the audio file.
   Thanks

[acrawford73]

sorry already mentioned