VASA-1: Lifelike audio-driven talking faces generated in real time • Hacker News Board

Re: VASA-1: Lifelike audio-driven talking faces generated in real time

#11

Post by fluffet » Thu, Apr 18, 2024, 1:08 PM UTC

This is absolutely crazy. And it'll only get better from here. Imagine "VASA-9" or whatever.

I thought deepfakes were still quite a bit away but after this I will have to be way more careful online. It's not far from behind something that can show up in your "YouTube shorts" feed and trick you if you didn't already know it was AI.

Re: VASA-1: Lifelike audio-driven talking faces generated in real time

#12

Post by gedy » Thu, Apr 18, 2024, 2:10 PM UTC

My first thought was "oh no the interview fakes", but then I realized - what if they just kept using the face? Would I care?

Re: VASA-1: Lifelike audio-driven talking faces generated in real time

#13

Post by feyman_r » Thu, Apr 18, 2024, 2:35 PM UTC

nycdatasci wrote: post #8

“We have no plans to release an online demo, API, product, additional implementation details, or any related offerings until we are certain that the technology will be used responsibly and in accordance with proper regulations.”

/s it doesn’t have the phrase LLM in the title

Re: VASA-1: Lifelike audio-driven talking faces generated in real time

#14

Post by justinclift » Thu, Apr 18, 2024, 2:44 PM UTC

nycdatasci wrote: post #8

“We have no plans to release an online demo, API, product, additional implementation details, or any related offerings until we are certain that the technology will be used responsibly and in accordance with proper regulations.”

> until we are certain that the technology will be used responsibly ...

That's basically "never" then, so we'll see how long they hold out.

Scammers are already using the existing voice/image/video generation apparently fairly successfully. :(

Re: VASA-1: Lifelike audio-driven talking faces generated in real time

#15

Post by darig » Thu, Apr 18, 2024, 2:49 PM UTC

acidburnNSA wrote: post #3

Oh no. "Cameras on please!" will be replaced by "AI generated faces off please!" in teams.

[dead]

Re: VASA-1: Lifelike audio-driven talking faces generated in real time

#16

Post by pxoe » Thu, Apr 18, 2024, 4:17 PM UTC

maybe making a webpage with 27 videos isn't the greatest web design idea

Re: VASA-1: Lifelike audio-driven talking faces generated in real time

#17

Post by sitzkrieg » Thu, Apr 18, 2024, 4:34 PM UTC

nycdatasci wrote: post #8

“We have no plans to release an online demo, API, product, additional implementation details, or any related offerings until we are certain that the technology will be used responsibly and in accordance with proper regulations.”

money will change that

Re: VASA-1: Lifelike audio-driven talking faces generated in real time

#18

Post by sitzkrieg » Thu, Apr 18, 2024, 4:35 PM UTC

pxoe wrote: post #16

maybe making a webpage with 27 videos isn't the greatest web design idea

the busted two scrolling sections on mobile really doesnt help

Re: VASA-1: Lifelike audio-driven talking faces generated in real time

#19

Post by smusamashah » Thu, Apr 18, 2024, 5:45 PM UTC

fluffet wrote: post #11

This is absolutely crazy. And it'll only get better from here. Imagine "VASA-9" or whatever. I thought deepfakes were still quite a bit away but after this I will have to be way more careful online. It's not far from behind something that can show up in your "YouTube shorts" feed and trick you if you didn't already know it was AI.

This is good but nowhere as good as EMO https://humanaigc.github.io/emote-portrait-alive/ (https://news.ycombinator.com/item?id=39533326)

This one has too much movement and looks eerie/robotic/uncanny valley. While EMO looks just perfect.

Re: VASA-1: Lifelike audio-driven talking faces generated in real time

#20

Post by ilaksh » Thu, Apr 18, 2024, 6:40 PM UTC

justinclift wrote: post #14

nycdatasci wrote: post #8

“We have no plans to release an online demo, API, product, additional implementation details, or any related offerings until we are certain that the technology will be used responsibly and in accordance with proper regulations.”

> until we are certain that the technology will be used responsibly ... That's basically "never" then, so we'll see how long they hold out. Scammers are already using the existing voice/image/video generation apparently fairly successfully. :(

Eventually someone will implement one of these really good recent ones as open source and then it will be on replicate etc. right now the open source ones like SadTalker and Video Retalking are not live and are unconvincing.