New Software Fakes Video Of Former US President Barack Obama

What will we do when the video clips we rely on for news are routinely and effortlessly faked – when it becomes nearly impossible to differentiate genuine statements from malicious imitations without a thorough investigation? Will you continue to believe every word uttered by US media organizations like CNN? Or will you slowly learn to question every piece of information you receive?

We’ve reported in the past about studies showing how effortlessly even a novice hacker could learn to edit videos of people speaking to make it seem as if their words were coming out of the mouth of a famous celebrity like, say, former US President Barack Obama.

Last year, a paper by the Stanford lab of Matthias Niessner titled "Face2Face: Real-time Face Capture and Reenactment of RGB Videos” shows how disturbingly easy it is to take a surrogate actor and, in real time, using everyday available tools, create the illusion that someone else, notably someone famous or important, is speaking. Even more disturbing: one doesn't need sophisticated equipment to create a "talking" clone - a commodity webcam and some software is all one needs to create the greatest of sensory manipulations.

And now, the Telegraph is reporting that a computer program has been created that can edit videos of people speaking to realistically make it look like they said something else, raising fears of clips being tampered with online.

As a reminder, here’s an excerpt from the Stanford paper:

We present a novel approach for real-time facial reenactment of a monocular target video sequence (e.g., Youtube video). The source sequence is also a monocular video stream, captured live with a commodity webcam. Our goal is to animate the facial expressions of the target video by a source actor and re-render the manipulated output video in a photo-realistic fashion. To this end, we first address the under-constrained problem of facial identity recovery from monocular video by non-rigid model-based bundling. At run time, we track facial expressions of both source and target video using a dense photometric consistency measure. Reenactment is then achieved by fast and efficient deformation transfer between source and target. The mouth interior that best matches the re-targeted expression is retrieved from the target sequence and warped to produce an accurate fit. Finally, we convincingly re-render the synthesized target face on top of the corresponding video stream such that it seamlessly blends with the real-world illumination. We demonstrate our method in a live setup, where Youtube videos are reenacted in real time.

In simple English: famous "talking heads" can be practically anyone masquerading as said celebrity, and due to the real time conversion, they can talk, react, answer questions and generally emote so that the deception is flawless and totally convincing.

Now, researchers at the University of Washington have lip-synced a video of former US President Barack Obama using the program to superimpose new audio onto the clip. The realistic results put words in Obama's mouth by converting audio sounds into mouth movements and blending them onto an existing video of speech.

Researchers say the software could be used in special effects, or to improve the quality of video calls.

"When you watch Skype or Google Hangouts, often the connection is stuttery and low-resolution and really unpleasant, but the audio is pretty good," said Steve Seitz, co-author of the research and professor at the University of Washington.


"So if you could use the audio to produce much higher-quality video, that would be terrific."

Until recently, video lip-syncing involved hours of filming and editing.  But the computer program can create a clip with new audio after analyzing one hour of speech rather than 14.  
So striking is the real time effect of the conversion, the creators of this algorithm felt the need to clarify their intentions:

This demo video is purely research-focused and we would like to clarify the goals and intent of our work. Our aim is to demonstrate the capabilities of modern computer vision and graphics technology, and convey it in an approachable and fun way. We want to emphasize that computer-generated videos have been part of feature-film movies for over 30 years. Virtually every high-end movie production contains a significant percentage of synthetically-generated content (from Lord of the Rings to Benjamin Button). These results are hard to distinguish from reality and it often goes unnoticed that the content is not real. The novelty and contribution of our work is that we can edit pre-recorded videos in real-time on a commodity PC. Please also note that our efforts include the detection of edits in video footage in order to verify a clip’s authenticity. For additional information, we refer to our project website (see above). Hopefully, you enjoyed watching our video, and we hope to provide a positive takeaway.

Stanford expands upon its methodology in the video below:


Mr 9x19 El Vaquero Fri, 07/14/2017 - 07:59 Permalink

it is fun technic for low fi fun for youtube, but when you see the render in a 720+ quality i garanty you can notice the image manipulation simply because of the texture materia around the nodes that are taken as reference points for anchoring. go install 3dsmax, apply any bitmap material on a sphere, set mesh mode,  displace one node, you will see what i am talking about. this technology is old school texture applying... obsolete. a reel progress is to set pure artificial multi layer rendered materia to prevent any deformation of the actual bitmaps.not gonna happen that soon. a full 1080 footage is still visually problematic, we have never been able to make realistic textures. avatar in BD is a joke for example.  

In reply to by El Vaquero

UselessEater peopledontwanttruth Thu, 07/13/2017 - 22:47 Permalink

it looks like this:… these losers got it all handed on a platter, life was waaaay better before but they were told they wanted a black leader/slaver and yet they complain and complain and decades after are still convinced it is all whitey fault even when the leaders and managers have been black for decades like in Zambia, Zimbabwe, Botswana...The tech is being used in many ways anf forms, if we're only hearing about it NOW its been around a long time and it was already used against us in many ways.Sick Sweden included, don;t let your jaw drop as you read some of this stuff...there is zero doubt they have not been subjected to tech like this - we do not naturally gravitate to suicide with rape and torture along the way.

In reply to by peopledontwanttruth

GUS100CORRINA UselessEater Thu, 07/13/2017 - 23:24 Permalink

One term: AUGMENTED REALITY!!In the hands of EVIL people, the persuasive power of this technology is unsettling. It becomes seductive and addictive like DRUGS and Rock;n Roll musicDefinition; Augmented RealityA technology that superimposes a computer-generated image on a user's view of the real world, thus providing a composite viewOBAMA's face as the image used would confirm that GOD's CURSE is on AMERICA.If the American people are willing to accept OBAMA in any future leadership role, then GOD has sent them strong delusion so that they will believe the lie and it is OVER!!!Remember HITLER when hearing people talk about how to use this technology in the political realm.

In reply to by UselessEater

overbet barndoor Thu, 07/13/2017 - 22:02 Permalink

Fuck that nagger.This is interesting:  “We all agree there is a need to provide technical details on Russian hacking. (Redacted) urges you and (redacted) to bring up again the issue with (redacted). It’s quite important, especially as we approach the hearings. If there are no technical details we have to find some by all means. I am sure his guys can do it.”…

In reply to by barndoor

Lore anonnn Fri, 07/14/2017 - 00:26 Permalink

Definitely not new.  I first saw this demonstrated with rendering of full body motion in real time at a computer lab in Western Canada back in Fall 2001.  They let us try it out.  When my turn came, I made a shiny render of C-3PO stand up from a seated position and wave. In the late 90s, I remember watching a national TV news announcer's face transform into a very crude neon-green polyhedral mesh typical of preliminary texture-mapping before the screen abruptly flicked back to the face of the live anchor - a glitch, presumably.I don't know much of the jargon, but there are different ways to spot fakes. Some have been popping up in online videos. It might be fun sometimes just to play 'spot the fraud.'

In reply to by anonnn

stinkypinky wisehiney Thu, 07/13/2017 - 22:21 Permalink

Those actually looked quite fake. Look carefully at the mouths - they look much more like video game approximations of speech than real. Compare to the left side true mouth movements. The only thing these people did well was to track the head movements and get the shapes and colors right; The actual mouth movements themselves are off.

In reply to by wisehiney

JuliaS stinkypinky Thu, 07/13/2017 - 22:37 Permalink

I say it's pretty good for something that didn't even exist yesterday. It will only get better. A year ago they could only do things in the second video, which involved some manual work. And it's gradually approaching full autopilot.The idea is to eventually let neural networks do all the fabrication. You feed the text and the computer will search the web for you, analyze facial features, speech patterns and fully construct the fake footage from scratch. The goal, of course, is to make this completely undetectable.Right now the animation is superimposed against real footage, so if you are in possession of the original, you can spot and expose false duplicates. If, on the other hand, a fabrication is all that exists, then it's a whole different story.Scary times dead ahead.

In reply to by stinkypinky

JuliaS stinkypinky Fri, 07/14/2017 - 00:33 Permalink

I work in film and television. This research is as old as computer graphics itself. Digital face replacement is all the rage when it comes to stunts, though such process typically doesn't involve lipsyncing. When full digital doubles are used they still look fake after all the years but that is mainly due to shrinking budgets and rigid time constraints. Every year you get less time to do a good job. That's why digitally revived Star Wars characters look worse than Benjamin Button from a decade ago. It's shot like a movie, but processed like episodic television.I've written a few algos myself, which were based on research done by Digital Emily Project team. They were used in a couple of movies, which I'm very proud of.

In reply to by stinkypinky

tmosley Thu, 07/13/2017 - 21:58 Permalink

I love how everyone has been telling me in these threads that AI is a pipe dream.Peanutz will always be behind the curve, by at LEAST two years. This becomes a problem in the age of AGI, where doublings happen at ever increasing rates.

a Smudge by an… tmosley Thu, 07/13/2017 - 23:38 Permalink

Not sure we're saying AI is a pipe dream, we're just saying it's not AI. It's sort of how people are surprised to see that some "algorithms" are in fact quite simple.

An AI is just like a business process flow chart. Let's say you have a business process that makes cars. The chart details the flow of car production from design to dealership and every step in between. And it can have reporting tools built in, it can have diagnostic tools built in, it can have any form of software and human inputs and raw materials control it requires. It can have "oh shit" scenarios to handle challenges. Whatever. But every step in that process was created by humans to obtain specific results. We can take this a step further like the Toyota production method which integrates quality control right back into design production and we're getting there.

An AI pretty much by definition is a software that models human interaction and response. And it's constructed as a freaking flowchart. Or "decision tree". And it employs tons of tools and sub processes to do this. Each one of which was modeled by a human with a specific objective.

Now it's fashionable to talk about "machine learning". This is variously interpreted. The Toyota production process can be said to have "machine learning" because machines are looking at quality control metrics OVER A PERIOD OF TIME and they can more quickly identify issues that humans might overlook.

facebook might be considered to have "machine learning" because there were so many theoretical behavioral modeling THEORIES out there but they didn't have nearly enough data to run the experiment. Well now they do and they have a considerable amount of TIME built up to there these models are a snap.

Based on this data (and the various sciences involved in coming up with responses to it), we now have PREDICTIVE CAPABILITY on human behavior. Who's likely to commit suicide today? How about a suicide bombing? Who's gonna be best in their 5th grade class?

None of these answers are gonna be definitive from a machine but they will all result in binary or % terms or probability factors or deltas or whatever. So that's the name of the game now. We just keep narrowing the statistical % of failure.

But is this machine learning? Not really so far. This is humans learning faster with machines.

Sorry if this is simplistic but I was really trying hard not to use words like "multispacial adaptive heuristics". Cause it's like one of my favorites.

In reply to by tmosley

ghengis86 Thu, 07/13/2017 - 22:34 Permalink

How many Osama videos have there been since 9/11? (((They've))) been doing this for a long time and with most tech, the plebs can get their hands on it a couple decades later.

indio007 Thu, 07/13/2017 - 22:02 Permalink

They fake our reality because we let it be faked.  Do you know how many people believe the garbage NASA puts out?They burn through $50 million a day so people can watch fake space.Shit people still believe NASA went to the moon in a homeless tweaker shelter.If your one of those people. Who panned the camera because I think they left someone behind? 

Anteater indio007 Thu, 07/13/2017 - 22:27 Permalink

$5 BILLION a year, with a 'B', still going to NASA for the RETIRED Space Shuttle Program,so that the Right Stuff can sit at their computer consoles and jibber-jabber with StevenHawkins, 'til they make their 30-and-out pensions-for-life. The Shuttle Program ended!That's $100B until the Right Stuff are dead, solar-cremated, in orbit, in a Space Blanket.

In reply to by indio007

pocomotion Thu, 07/13/2017 - 22:15 Permalink

Can someone do a video of me?  Yes I would like to be addressing the nation to tell them1.  You have been had.2.  You would not believe me but you will have babies from the blue helmet guys and not know the father.  In war it takes 25 guys grunting on top of you to give you that baby.3.  Let me talk a but on how we should handle the world bankers.4.  Deal with the Lawyers5.  Deal with the government, politicians, CEOs and CFOs6.  Outlaw SECRET SOCIETIES.7.  Print our own money.8.  Send Israel a Dear John letter.9.  Cancel ALL DEPT10.  Rebuild our military and have them totally adhere to the CONSTITUTION OF THE UNITED DATES. Thanks, that would be really neat.