News for hackers. Stuff that matters.
Re:A speech-to-text practitioner’s criticisms of industry an... by ragebol on Monday April 06, 2020 @04:22AM (#22791805)
I'm no way an expect in any of this, but I'd expect that there would be a lot of features common between a lot of languages, akin to the International Phonetic Alphabet [0]. Pre-training on all languages to get those shared features could make it easier to fine-tune eg. English on top perhaps. Or not, just pondering here.

[0] https://en.wikipedia.org/wiki/International_Phonetic_Alphabe...

Re:A speech-to-text practitioner’s criticisms of industry an... by tasogare on Monday April 06, 2020 @11:33AM (#22794604)
Good luck finding records of the 6000+ existing languages...
Re:A speech-to-text practitioner’s criticisms of industry an... by nshm on Monday April 06, 2020 @12:51PM (#22795319)
You can check https://github.com/festvox/datasets-CMU_Wilderness, it has recordings of 700 languages created from New Testaments from http://www.bible.is/