Hacker News Reader

As I understand it, all the voice assistants rely on a backend to do the voice processing, at a minimum.

There are several freely available voice datasets.

https://voice.mozilla.org/data

If everyone who reads HN commits a little time to Mozilla Common Voice in 2018, we might have a non-cloud solution in 2019.

This is awesome. Thank you for pointing me in this direction. This kind of corpus is going to be extremely important to getting a more diverse array of voice applications out there in the open.

Reply on news.ycombinator.com