Would it be difficult to make a program that sits in the audio stack (probably o... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

foota on July 5, 2017 | parent | context | favorite | on: Top accessibility complaints from people with disa...

Would it be difficult to make a program that sits in the audio stack (probably on Windows) and tries to run voice recognition on everything coming over the line and then renders anything that sounds like speech at the bottom of the screen?

Garfgon on July 5, 2017 | [–]

Yes. Youtube tries (tried?) to autocaption videos, and the results were infamously bad.

rhizome on July 5, 2017 | [–]

Yes. Voice recognition for arbitrary voices, accents, and lexicons is not as reliable as you imagine, much less in real-time.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact