Not yet a member? No problem!
Sign-up just takes a second.
Forgot your password?
Recover it now.
Already signed up?
Log in now.
Forgot your password?
Recover it now.
Not yet a member? No problem!
Sign-up just takes a second.
Remember your password?
Log in now.
14 Comments
deathcow*promote probably a dupe I cant find it
siftbotSelf promoting this video and sending it back into the queue for one more try; last queued Sunday, January 29th, 2017 10:55am PST - promote requested by original submitter deathcow.
mxxcon*related=https://videosift.com/video/Real-time-Expression-Transfer-for-Facial-Reenactment
siftbotReal-time Expression Transfer for Facial Reenactment has been added as a related post - related requested by mxxcon.
newtboyGreat....so now when producers edit audio to change what's actually been said, they can add video of the person saying the made up phrases. That was one way I determined if what I'm watching is real or edited....when they would splice audio together to create a monologue/dialogue, they had to cut away from the speakers face so you wouldn't notice the 1975 kungfu level dubbing....now they can convincingly fake both....in real time? We're totally fucked. Reality just became obsolete.
I'm going to Mars.
newtboyTo parrot myself above....*terrible
siftbotAdding video to channels (Terrible) - requested by newtboy.
cosmovitellisays...Oceania was at war with Eastasia: Oceania had always been at war with Eastasia.
PlayhousePalsThis is so *wtf
siftbotAdding video to channels (Wtf) - requested by PlayhousePals.
PaybackLulz, what is it with people doing shit because they can, and never asking if they should? I thought the real time voice mimicry tech was bad enough...
entr0pyI don't think this is the sort of thing that wouldn't be happening if not for Stanford, but rather it wouldn't be happening in a transparent and openly published way. The entertainment and gaming industries alone guarantee these techniques will make steady progress.
In the description they mention a big part of their work is the detection of video edits to verify the authenticity of videos. Trying to keep state of the art programming techniques in the public domain can only help us spot this stuff and be less gullible as a society.
Lulz, what is it with people doing shit because they can, and never asking if they should? I thought the real time voice mimicry tech was bad enough...
LiquidDriftsays...If people want to believe it, it won't matter if you tell them it's a fake video.
I don't think this is the sort of thing that wouldn't be happening if not for Stanford, but rather it wouldn't be happening in a transparent and openly published way. The entertainment and gaming industries alone guarantee these techniques will make steady progress.
In the description they mention a big part of their work is the detection of video edits to verify the authenticity of videos. Trying to keep state of the art programming techniques in the public domain can only help us spot this stuff and be less gullible as a society.
siftbotA sneak peek of the future of Fake News: Lip Synching has been added as a related post - related requested by bareboards2 on that post.
Discuss...
Enable JavaScript to submit a comment.