Researchers Have Figured Out How To Fake News Video With AI (qz.com)

An anonymous reader quotes a report from Quartz: A team of computer scientists at the University of Washington have used artificial intelligence to render visually convincing videos of Barack Obama saying things he's said before, but in a totally new context. In a paper published this month, the researchers explained their methodology: Using a neural network trained on 17 hours of footage of the former U.S. president's weekly addresses, they were able to generate mouth shapes from arbitrary audio clips of Obama's voice. The shapes were then textured to photorealistic quality and overlaid onto Obama's face in a different "target" video. Finally, the researchers retimed the target video to move Obama's body naturally to the rhythm of the new audio track. In their paper, the researchers pointed to several practical applications of being able to generate high quality video from audio, including helping hearing-impaired people lip-read audio during a phone call or creating realistic digital characters in the film and gaming industries. But the more disturbing consequence of such a technology is its potential to proliferate video-based fake news. Though the researchers used only real audio for the study, they were able to skip and reorder Obama's sentences seamlessly and even use audio from an Obama impersonator to achieve near-perfect results. The rapid advancement of voice-synthesis software also provides easy, off-the-shelf solutions for compelling, falsified audio. You can view the demo here: "Synthesizing Obama: Learning Lib Sync from Audio"

Comments

Fake Research

[Anonymous Coward]

"You didn't build that."

Foreigners says no.

[Zaatxe]

I grew up in a country where most foreign shows and movies are shown dubbed on TV. The result videos look exactly like a dubbed movie to me. My brain automatically says "this is not the original audio from this footage" when watching this.

Block chain to the rescue?

[Aristos Mazer]

Sci-fi author Ian Banks had this problem in his futuristic The Culture novels. His solution: have repositories of record scattered everywhere that any recording device would upload to in real time. The repository would cryptographically sign the recording then download it back to the recording device. That provided both the sequence ordering and verifiability for the recordings at any future date. Banks wrote that before the arrival of block chain tech. The crowdsourced signing ability of the block chain is being used for things far beyond digital currency already. Seems like we need recording devices that can add hashes of their recordings to the block chain so that there is a record of where and when a given video was shot that cannot be falsified or denied. If we get to the point where most commercial recorders are using that service, we could once again have verifiable news. Seem viable?