AWS
AWS offers reliable, scalable, and inexpensive cloud.
Start new thread

Amazon Polly - Turn text into lifelike speech using deep learning

by•

Add a comment

Replies

Best
i still find the voices to sound artificial, could be ages before they really sound natural
Wow, this is really cool. It could, for example, enable voice customer support where the support rep is typing, but users computer is speaking it out. Intercom dudes, free idea for ya :p
Why is it that the female voices sound so much more realistic than the male voices?
in the US it would seem we have gotten accustomed to generic female voices (unlike for example Germany re: GPS). but there's actually more psychology here, see re: the genericity and likeability of women's vs men's voices
Spanish, Danish both sound the same to me. Male voice is clearly robotic.
Hunter
Amazon Polly is a service that turns text into lifelike speech. Polly lets you create applications that talk, enabling you to build entirely new categories of speech-enabled products.
Pocket should implement with this. I'll pay a couple bucks per month for premium text to speech
I've implemented this and have it working locally; reads your Pocket, turns it into a podcast. Problem currently is that the cheap voices (eg Amazon Polly) are too robotic and the good voices (Google WaveNet) are way too expensive. ($16 per MB of text, ~$0.10+ per article)
Developed a Free Text to Speech Web App based on Amazon Polly :