An open source implementation of DeepVoice 3: 2000-Speaker Neural Text-to-Speech

Re:An open source implementation of DeepVoice 3: 2000-Speake... by tekkk on Thursday April 26, 2018 @06:23AM (#16930571)

Recently I started digging into text-to-speech implementations of which Google seems to have the best know-how.

https://google.github.io/tacotron/publications/global_style_...

Those samples sound already frighteningly human.

Not to disparage the DeepVoice3 in anyway! There must have been a lot of work put into it.

[ Parent | on HN ]

Re:An open source implementation of DeepVoice 3: 2000-Speake... by throwaway84742 on Thursday April 26, 2018 @06:54AM (#16930736)

This is scary good. The only way you can tell these aren’t real is because they don’t (yet) model the lung capacity, so some phrases are longer than a human would be able to pronounce on one breath. I can’t wait for this stuff to become mainstream.

[ Parent | on HN ]

Re:An open source implementation of DeepVoice 3: 2000-Speake... by romaniv on Thursday April 26, 2018 @01:23PM (#16933795)

>Google seems to have the best know-how.

Funny how they publish their "research papers", yet no one else is able to implement their engine with even remotely comparable results.

[ Parent | on HN ]

Re:An open source implementation of DeepVoice 3: 2000-Speake... by mkagenius on Thursday April 26, 2018 @01:43PM (#16933953)

Oh, is it not verifiable?

[ Parent | on HN ]

Re:An open source implementation of DeepVoice 3: 2000-Speake... by heywire on Thursday April 26, 2018 @02:33PM (#16934358)

Wait, are you saying that the very first audio sample on that page is NOT a human?

[ Parent | on HN ]