In the 11 years I’ve been writing this blog, very few topics have touched a nerve with the VO community like the prospect of synthetic voices overtaking our turf.

We naturally (and maybe rightfully) believe that no digital formula…no artificial intelligence can EVER capture the nuance of human speech well enough to really have an impact on our bottom line.

...the latest research from Google has reached a level of approximation of human vocal inflection that is quite honestly threatening... Click To Tweet

That hasn’t kept many a geek from trying.  And they’re still at it.  In May of last year, I wrote about Adobe Labs working on VOCo .

Today, I’m back to tell you that the latest research from Google has reached a level of approximation of human vocal inflection that is quite honestly threatening.

They’ve even named their process: PROSODY – “…the collection of expressive factors of speech, such as intonation, stress, and rhythm…”

Still an unbeliever?  Then take a stroll on over to their publication about Prosody, and listen to their samples.

Scary good.

Dave? ( a reference to 2001, A Space Odyssey)

Here’s the link:

https://research.googleblog.com/2018/03/expressive-speech-synthesis-with.html

CourVO

Comments

comments