Thursday, October 26, 2017

Deep Voice 3: Neural Network Text-to-Speech - 3 audio samples: 1 speaker (~20 hours of data) Versus 108 speakers (~44 hours) Versus 2484 speakers (~820 hours)

http://ift.tt/2yNgjcx

Submitted October 26, 2017 at 11:46AM by bboyjkang http://ift.tt/2gEivIX via TikTokTikk

No comments:

Post a Comment