Amazon Polly – Lifelike Text-To-Speech • Hacker News Board

Re: Amazon Polly – Lifelike Text-To-Speech

#41

Post by splike » Wed, Dec 07, 2016, 4:40 PM UTC

mwcampbell wrote: post #40

splike wrote: post #36

Earlier quoted context omitted.

I agree that the voice sounds pretty bad, but I think the 'lifelike' part comes from the intelligence part. I.E WA -> Washington, and 75F -> 75 Fahrenheit

Text-to-speech systems have been expanding abbreviations for decades, and it always backfires in some cases. For example, DECtalk would expand "Sun" to "Sunday" regardless of context, with the humorous result that blind people reading tech news in the 90s would often hear about Sunday Microsystems. Edit: Yes, that example was unfair, because it's from the 90s (and actually, DECtalk was largely unchanged since the 80s…

Hence the need for intelligence

Re: Amazon Polly – Lifelike Text-To-Speech

#42

Post by angstrom » Wed, Dec 07, 2016, 4:40 PM UTC

hacker_9 wrote: post #25

derefr wrote: post #13

One thing I always still notice about these "lifelike" speech models is that they still have random pitch variation that wouldn't be present from a real speaker—sort of a "warbling", similar to audio heavily compressed by some cellular realtime audio codecs. I assume this is due to random variation in the pitches the speakers in the training data used to say given phonemes. Would this, then, imply that speech models…

Additionally when we speak we don't just 'replay' words; we add emotions and so on. Reminds me of this video [1] analysing how the actor Anthony Hopkins converts his lines to speech for his scenes. [1] https://www.youtube.com/watch?v=4kSGkGKwp9U

The spoken word has so many nuances. Cadence, inflexion, tone, and then the facial performances are never captured by pure text to speech. You'd almost have to be able to set boundaries for seriousness, playfulness, innuendo, indifference, sarcasm etc because none of that is conveyed in the literal text without additional descriptions or contextual analysis of the conversation.

Re: Amazon Polly – Lifelike Text-To-Speech

#43

Post by IshKebab » Wed, Dec 07, 2016, 4:42 PM UTC

TazeTSchnitzel wrote: post #16

Could we drop “Lifelike” from the title? It's marketing puffery (and untrue, it sounded very robotic to me to the point I was convinced I'd heard the wrong audio file). The actual story is that it's Amazon's text-to-speech-as-a-service offering, not that it's particularly innovative in terms of sounding good.

Sooo cynical. This is way better than most TTS systems, including Google's. The only one I've ever heard that is better is DeepWave, and that suffers from lots of background noise.

Re: Amazon Polly – Lifelike Text-To-Speech

#44

Post by rckclmbr » Wed, Dec 07, 2016, 4:45 PM UTC

derefr wrote: post #13

One thing I always still notice about these "lifelike" speech models is that they still have random pitch variation that wouldn't be present from a real speaker—sort of a "warbling", similar to audio heavily compressed by some cellular realtime audio codecs. I assume this is due to random variation in the pitches the speakers in the training data used to say given phonemes. Would this, then, imply that speech models…

Southerns have an accent. New Jerseyans have an accent. British have an accent. Can't computers have an accent, too?

Re: Amazon Polly – Lifelike Text-To-Speech

#45

Post by derefr » Wed, Dec 07, 2016, 4:46 PM UTC

iainmerrick wrote: post #18

derefr wrote: post #13

One thing I always still notice about these "lifelike" speech models is that they still have random pitch variation that wouldn't be present from a real speaker—sort of a "warbling", similar to audio heavily compressed by some cellular realtime audio codecs. I assume this is due to random variation in the pitches the speakers in the training data used to say given phonemes. Would this, then, imply that speech models…

Sheer speculation, but I wonder if that could be related to the "checkerboarding" artifacts you almost always see in AI-generated images? Here's a good writeup: http://distill.pub/2016/deconv-checkerboard/

Re: the checker-boarding artifacts, that sounds interestingly like deconvolutional NNs treat their samples, at each deconv layer, as independent, rather than taking the Rule of Addition of probability-weights (P(A ∪ B) = P(A) + P(B) - P(A ∩ B)) into account. (Would a model that did do this then be a "Bayesian Deconvolutional Neural Network"? Does that even make sense?)

I doubt this would result in audible artifacts at the audio level, though. The visual checkerboarding appears at the 1px detail-level of the output image—this would translate more to minor fluctuations along an audio sample's Nyquist frequency, which is effectively undetectable if the output sampling rate is of the usual kind (44.1kHz).

Re: Amazon Polly – Lifelike Text-To-Speech

#46

Post by shiftpgdn » Wed, Dec 07, 2016, 4:46 PM UTC

ckluis wrote: post #14

sagivo wrote: post #8

#offtopic a bit but i wonder if anyone knows a good api for the other-way around -> speech to text

Shoot - I'd love an API you pass in txt in one language and get multiple languages back in txt and speech.

Amazon Lex and Google's GCE offer this.

Re: Amazon Polly – Lifelike Text-To-Speech

#47

Post by zackbloom » Wed, Dec 07, 2016, 4:46 PM UTC

Here's a sound clip of one of the voices I pulled, if you'd like to hear it: https://soundcloud.com/zack-bloom/amazons-new-text-to-speech...

Re: Amazon Polly – Lifelike Text-To-Speech

#48

Post by aninhumer » Wed, Dec 07, 2016, 4:46 PM UTC

aikinai wrote: post #20

derefr wrote: post #13

One thing I always still notice about these "lifelike" speech models is that they still have random pitch variation that wouldn't be present from a real speaker—sort of a "warbling", similar to audio heavily compressed by some cellular realtime audio codecs. I assume this is due to random variation in the pitches the speakers in the training data used to say given phonemes. Would this, then, imply that speech models…

I'm not sure about Mandarin, but this is definitely true for Japanese. Near-human Japanese text-to-speech has been around a very long time.

True, but Japanese isn't really a tonal language, so it doesn't really evidence the parent's point.

Re: Amazon Polly – Lifelike Text-To-Speech

#49

Post by jeena » Wed, Dec 07, 2016, 4:47 PM UTC

I really like text2speech because I really like the computer to read articles to me while I'm doing something else, but I won't use any internet service for that just to get a little bit more quality in comperison to Apples "Alex" or MeryTTS http://mary.dfki.de/

Re: Amazon Polly – Lifelike Text-To-Speech

#50

Post by dharma1 » Wed, Dec 07, 2016, 4:48 PM UTC

IshKebab wrote: post #43

TazeTSchnitzel wrote: post #16

Could we drop “Lifelike” from the title? It's marketing puffery (and untrue, it sounded very robotic to me to the point I was convinced I'd heard the wrong audio file). The actual story is that it's Amazon's text-to-speech-as-a-service offering, not that it's particularly innovative in terms of sounding good.

Sooo cynical. This is way better than most TTS systems, including Google's. The only one I've ever heard that is better is DeepWave, and that suffers from lots of background noise.

Do you mean WaveNet? The noisiness is because it uses 8bit (mu-law) audio. 8bit is because WaveNet is computationally extremely expensive. It's pretty much proof of concept at this stage, not production ready TTS.