>This post presents WaveNet, a deep generative model of raw audio waveforms. We show that WaveNets are able to generate speech which mimics any human voice and which sounds more natural than the best existing Text-to-Speech systems, reducing the gap with human performance by over 50%.
>We also demonstrate that the same network can be used to synthesize other audio signals such as music, and present some striking samples of automatically generated piano pieces.
Yes the quality is close to human sound but the point is seiyuu is a huge part of anime/J-game industry financially. It's idolized already.
You can create another Hatsune Miku sure but it's just not comparable to the whole seiyuu industry money wise.
So large companies won't use it but the real benefit of such technology is that individuals and amateur animation group can create animation easier without worrying about having to get shitty cheap voice actors anymore.
Ethan Ross
>Seiyuu will be replaced by computers It doesn't seem you know much of the Japanese VA industry.
Jace Perry
Cool idea. Wake me up in 2 decades when it matters.
Zachary Green
>So large companies won't use it but the real benefit of such technology is that individuals and amateur animation group can create animation easier without worrying about having to get shitty cheap voice actors anymore.
It's more likely that it will take off in video games first. Realistic voice synthesis would allow developers to produce games with a whole lot more dialog.
Xavier Gonzalez
I look forward to it, hopefully they'll use it for actual robot characters
Jack Perry
It could actually be a boon for western dubbing.
Imagine your favorite animu characters speaking in English, but still sounding like the original nihonjin voice actors.
But that's probably still a thing for years, perhaps even decades into the future.
Dylan Ortiz
I'm ready for the computerized fandubs of the future
Oliver Morales
>Imagine your favorite animu characters speaking in English, but still sounding like the original nihonjin voice actors.
Even better, imagine the dubbed voices not sounding like total crap? Robots could hardly be worse.
Christopher Cook
THUGGERY
Austin Brooks
GLorious WAIFU AGE!!!!!!!!
Nicholas Sullivan
The piano part was not that good.
Nicholas Edwards
>Computers playing the piano Spooky
Hunter Parker
Will it sing daisy bell?
Andrew Lee
>Seiyuu will be replaced by computers They said the same thing about Vocaloids.
Aaron Rivera
Vocaloids will be replaced by seiyuus.
Adam Lee
But user, vocaloids are based on seiyuus.
Asher Allen
Shocking revelation.
Angel Ortiz
>no sample
Must be hard being fucking retarded.
Lucas Taylor
Cthuko shall live again.
Oliver White
In twenty years we'll have perfect copies of voice actors able to say whatever you want in whatever language you want and Sup Forums will still prefer subtitles.
Grayson Cox
Yes. The future of eroge is glorious.
Xavier Cooper
I doubt it. Most anime producers aren't interested in fucking a computer.
Benjamin Walker
So does it mean I will have voiced waifu AI in the future?
Evan Scott
>perfect copies You might have learned how languages work by then though.
Chase Anderson
Yes, but not just that. It will be your waifu's voice specifically, since the network will have learned how to sound just like her.
Justin Reed
It's not right.
Michael James
Neat. >tfw your waifu will wake you up >tfw you can freely have a communication with her Just hope we will get artificial-personality technology soon and it would be perfect.
Evan Ward
Make it train more piano song samples and it'll get good user.