search

marktnoonan / transcription

Live Transcription based on Speech Recognition API
https://freelivetranscript.com
MIT License
35 stars 22 forks source link

Future features #4

Open marktnoonan opened 6 years ago

marktnoonan commented 6 years ago

  1. "Artsy" fork that does cool/crazy effects based on various things about what person is saying, how quickly they talk, whatever? Kind of like a winamp visualizer for a slam poetry reading. Tony had this idea on the way out the door.

  2. "Speech accuracy" version? I know some people who are trying to speak more clearly because they have a disability. If there was a way to use this speech recognition API to give them positive feedback for saying particular words well, it could be a neat way for them to independently work on pronouncing things.

  3. "Subtitle" style version, where it's just 2 lines of text at the bottom of something... like maybe a chrome extension that could put the subtitles at the bottom of the screen over a slideshow? Seems like there could be a lot of wrinkles with the condensed space... but the words would be captured by any screen recordings or webcasts of the content, so that's pretty neat.

All of this seems to suggest there ought to be one central piece of code that just does "I listen to audio and I give you back words", and if that's solid, then we can marry it to completely unrelated front end projects.

marktnoonan commented 6 years ago

2 could really be thought of as "guitar hero for words" or something like those typing tutors/games