Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

waxy.org

Post reply on HN
Print view View original

181–190 of 207 posts

Re: Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

#181

Post by lhl »

modeless wrote: post #158
lhl wrote: post #23

Earlier quoted context omitted.

For neutral sounding very fast/efficient voices, I find Coqui TTS VITS models to be very good. For slower, more expressive voice or voice cloning I think the Coqui TTS XTTS is good (or you can look at the mrq/tortoise-tts). I'm still awaiting a StyleTTS2 implementation. The audio samples sound top notch: https://styletts2.github.io/

You're in luck, the code dropped 6 hours ago :) https://github.com/yl4579/StyleTTS2 Looks promising, I'm going to check it out too! MIT license, even! If it's fast enough for real time, it could be the new best option. The paper claims faster inference than VITS...

Ha awesome! I just checked the repo literally before I posted and it was still empty, thanks for the heads up, will give it a spin now.

Re: Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

#182

Post by singleshot_ »

sofixa wrote: post #168
flextheruler wrote: post #164

Earlier quoted context omitted.

Your government class didn’t cover common law versus case law?

You probably mean common law, also sometimes known as case law, vs civil law which traces it's origins to the Napoleonic civil code, and which is used in all of the world outside of the former British colonies. My law classes did cover common law, yes, but not favourably(can you guess I come from a civil law country?). Sounds like a system that made sense in 15th century Britain, but is quite the complex beast with m…

> Sounds like a system that made sense in 15th century Britain

Eleventh.

Re: Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

#183

Post by JohnFen »

Legend2440 wrote: post #120
JohnFen wrote: post #119

Earlier quoted context omitted.

Which are, in my view, really minor advantages when compared to the disadvantages. Not only in terms of putting people out of work, but in terms of increasing the artifice of the world around us and decreasing its humanity.

"putting people out of work" by automating jobs is also a good thing. The amount of stuff humans can accomplish is strongly limited by the supply of workers. Automating one job frees them up to do other things.

> "putting people out of work" by automating jobs is also a good thing

Unless you're one of the people out of work. And even if you don't care anything about them, if there's enough of them then the resulting unrest will be your problem anyway.

Re: Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

#184

Post by JohnFen »

cm2012 wrote: post #137
JohnFen wrote: post #119

Earlier quoted context omitted.

Which are, in my view, really minor advantages when compared to the disadvantages. Not only in terms of putting people out of work, but in terms of increasing the artifice of the world around us and decreasing its humanity.

There's little nothing more important to the happiness of humanity than increased productivity per capita. That sounds crazy but when you think about it it's true.

> That sounds crazy but when you think about it it's true.

I've thought a lot about it, and I don't think it's true.

Re: Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

#185

Post by lhl »

lhl wrote: post #181
modeless wrote: post #158

Earlier quoted context omitted.

You're in luck, the code dropped 6 hours ago :) https://github.com/yl4579/StyleTTS2 Looks promising, I'm going to check it out too! MIT license, even! If it's fast enough for real time, it could be the new best option. The paper claims faster inference than VITS...

Ha awesome! I just checked the repo literally before I posted and it was still empty, thanks for the heads up, will give it a spin now.

Just a followup for those interested, inference implementation notes and comparison clip between StyleTTS2, TTS VITS, and XTTS: https://fediverse.randomfoo.net/notice/AaOgprU715gcT5GrZ2

Re: Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

#186

Post by modeless »

lhl wrote: post #185
lhl wrote: post #181

Earlier quoted context omitted.

Ha awesome! I just checked the repo literally before I posted and it was still empty, thanks for the heads up, will give it a spin now.

Just a followup for those interested, inference implementation notes and comparison clip between StyleTTS2, TTS VITS, and XTTS: https://fediverse.randomfoo.net/notice/AaOgprU715gcT5GrZ2

Wow you got it working so fast! I'm still stuck in package manager hell trying to debug a million little issues.

Re: Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

#187

Post by colechristensen »

azinman2 wrote: post #174
colechristensen wrote: post #12

Earlier quoted context omitted.

Indeed. If you're not using a cloud provider bandwidth is extremely cheap. In fact locally I can get a 10 gbps home internet unmetered connection for $300/mo. I'm not sure how they'd react if I transferred 1 PB/mo though :)

That’s pretty expensive. Sonic offers 1-10gbps (depending on where you live) unmetered symmetric connections for $60/mo to the Bay Area… they’re also the only ISP that petitioned the FCC in favor of net neutrality. For work I end up transferring 50-150 gigs often, sometimes daily. Never heard a word from them that this has been a problem.

That's pretty cool, but I'd say the opposite that Sonic is crazy cheap.

Re: Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

#188

Post by colechristensen »

tomrod wrote: post #96
colechristensen wrote: post #12

Earlier quoted context omitted.

Indeed. If you're not using a cloud provider bandwidth is extremely cheap. In fact locally I can get a 10 gbps home internet unmetered connection for $300/mo. I'm not sure how they'd react if I transferred 1 PB/mo though :)

Is my math wrong here? 10 gbps -> 8s per 10 GB -> 800s per 1TB -> 80,000s per 1PB -> 22.3 hrs at full speed for 1 PB?

Fully saturated you could transfer a few petabytes per month on a 10gig line.

Re: Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

#189

Post by gwern »

techdragon wrote: post #161
Legend2440 wrote: post #113

Earlier quoted context omitted.

They don't need that - they already have enough data to generate plausibly human voices that don't sound like anyone in particular. Voice cloning is a special case, these models are equally good at making new voices.

I’ve found it’s not actually as easy to get this stuff to sound different to the specific someone it’s trained on.

Don't expect that to last more than a year or two, assuming it's even still a problem for the best voice-generation AIs. Generating high-quality is the hard problem; generating specific high-quality samples is, by comparison, a lot easier.

Remember when Stable Diffusion was released a year ago and one of the big artist copes was "sure, it can generate random images, but it'll never be able to generate the same character repeatedly!" They were already wrong because Textual Inversion and DreamBooth were already published, and soon enough, ported to SD and now people could dump out thousands of images of the same character in the same consistent style etc (and did).

Re: Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning

#190

Post by techdragon »

gwern wrote: post #189
techdragon wrote: post #161

Earlier quoted context omitted.

I’ve found it’s not actually as easy to get this stuff to sound different to the specific someone it’s trained on.

Don't expect that to last more than a year or two, assuming it's even still a problem for the best voice-generation AIs. Generating high-quality is the hard problem; generating specific high-quality samples is, by comparison, a lot easier. Remember when Stable Diffusion was released a year ago and one of the big artist copes was "sure, it can generate random images, but it'll never be able to generate the same charac…

The issue is more that I can’t get the equivalent of a slider control to adjust one or more properties of the voice from the AI in real time… like a vocal fry slider to use an example of something most people are capable of deliberately doing when they want to… but the currently available models are pre-trained to sound like the average/median of one specific person (or character) and while I imagine tools will improve to control and customise the training of the models to customise this vocal output I don’t see a clear path from the current model architectural design to one where I can freely control the stylistic expression aspects of the vocal output without loading in a completely different set of model data trained for that new desired output.
Post reply on HN