News for hackers. Stuff that matters.

A comment on Voice2json: Offline speech and intent recognition on Linux. Parent comment

Re:Voice2json: Offline speech and intent recognition on Linux by sodality2 on Friday May 21, 2021 @02:02PM (#27237584)
> Imagine someone like Mozilla trying to collect hundreds of voice clips from each of tens of millions of users!!

They do, and it's working! https://commonvoice.mozilla.org/en

[ Parent | on HN ]
Re:Voice2json: Offline speech and intent recognition on Linux by londons_explore on Friday May 21, 2021 @02:34PM (#27238040)
Except they have 12k hours of audio, when really they could do with 12B hours of audio...
[ Parent | on HN ]
Re:Voice2json: Offline speech and intent recognition on Linux by woodson on Friday May 21, 2021 @03:57PM (#27239136)
Then you need a lot of people that listen to those 12B hours of audio, and multiple listeners agree for each chunk of audio that what is spoken corresponds to the transcript.
[ Parent | on HN ]
Re:Voice2json: Offline speech and intent recognition on Linux by londons_explore on Friday May 21, 2021 @05:51PM (#27240420)
Lots of machine learning systems can use unsupervised and semi-supervised learning. Then nobody has to listen to and annotate all that audio.
[ Parent | on HN ]
Re:Voice2json: Offline speech and intent recognition on Linux by woodson on Saturday May 22, 2021 @05:18PM (#27250353)
Yes, but then you don't need Mozilla collecting read speech samples. You can just scrape any audio out there, run speech activity detection, and there you go.
[ Parent | on HN ]
Re:Voice2json: Offline speech and intent recognition on Linux by sodality2 on Friday May 21, 2021 @02:36PM (#27238068)
Good point. I'm doing my part to contribute to it, though, not much else I can do!
[ Parent | on HN ]
Re:Voice2json: Offline speech and intent recognition on Linux by totetsu on Saturday May 22, 2021 @11:15AM (#27247325)
Didn't Moz fire all the people who worked on this, and they have started their own project now?
[ Parent | on HN ]