SoylentNews is people
SoylentNews
There has been some controversy over Deepfakes, a process of substituting faces in video. Almost immediately, it was used for pornography. While celebrities were generally unamused, porn stars were alarmed by the further commodification of their rôle. The algorithm is widely available and several web sites removed objectionable examples. You know something is controversial when porn sites remove it. Reddit was central for Deepfakes/FakeApp tech support and took drastic action to remove discussion after it started to become synonymous with fictitious revenge porn and other variants of anti-social practices.
I found a good description of the deepfakes algorithm. It runs via a standard neural network library but requires considerable processing power on specific GPUs. I will describe the video input (with face to be removed) as the source and the face to be replaced as the target. The neural network is trained with the target face only. The source is distorted and the neural network is trained to approximate reference images of the target. When the neural network is given the source, it has been trained to "undistort" the source to target.
[Continues...]
If there are multiple faces in a frame of video, face recognition restricts input to the most likely face. Indeed, for maximum efficiency, this technique is used to crop source video in all cases. The trick that makes the process feasible is that the neural network is only trained with the target face. Furthermore, given the use of libraries, the unique code to achieve this objective is shockingly small.
A friend attempted to mix DeepFakes with the Internet meme of Downfall parodies. There is an infamous scene in the film Downfall (not to be confused with the film Falling Down) where Adolf Hitler rants prior to defeat. Unfaithful subtitles of the German dialog have been used to parody everything from corporate sales targets to sportsball management to the ongoing medical abuse of transsexual patients. Until now, only the words in the subtitles changed. The audio and video was otherwise unchanged. My friend hoped that it would be possible to insert the likeness of people being parodied.
Unfortunately, it doesn't work with the current algorithm. The number of faces is not a problem. The clipping and occlusion prevents the neural network from working effectively. It should be possible with an extension of the current algorithm but it is currently impractical.
A further development, found by the same friend, is the automatic conversion of a one sentence description into a very short video. The example system uses Flintstones cartoons. An example sentence would be "Fred dancing in the kitchen" and a rough but valid video is created which matches the description. Potentially, it would be possible to automatically convert a novel into a 100 minute film with no human intervention. Given that novels are frequently converted into films, there is a large amount of example data which may be used as reference. I know this would only be moderately easier than making a holodeck but experts may not be aware of the progress towards either goal.
takyon: An algorithm can also be used to manipulate facial movements to match video or audio input (see this example of Jordan Peele controlling Barack Obama's face). DARPA is holding an event that will task experts with making and catching "deepfakes".
Researchers have also created short "movies" (64x64, 32-frame animated GIFs) from text descriptions. It may be possible to synthesize scenes for a full length movie in the future without needing strong AI. After all, procedural generation could be used to create and populate a virtual city (like the one in Big Hero 6), and then it's a matter of writing some kind of coherent narrative and "shooting" it. A "director neural network" could be trained to mimic the cinematography techniques of films created by humans, and then apply the results to the virtual environment.
(Score: 2) by MichaelDavidCrawford on Wednesday May 30 2018, @12:42AM (9 children)
The first engineer to actually accomplish this will put even Jeff Bezos completely to shame:
Hollywood actors can be... difficult... to work with.
Yes I Have No Bananas. [gofundme.com]
(Score: 3, Funny) by MostCynical on Wednesday May 30 2018, @01:00AM
Not as difficult when they are dead: http://www.vulture.com/2017/12/star-wars-the-last-jedi-carrie-fisher-death.html [vulture.com]
Or weren't alive in the first place:
http://starwars.wikia.com/wiki/List_of_changes_in_Star_Wars_re-releases [wikia.com]
"I guess once you start doubting, there's no end to it." -Batou, Ghost in the Shell: Stand Alone Complex
(Score: 4, Interesting) by takyon on Wednesday May 30 2018, @04:52AM (7 children)
We're within striking distance, but the animation is still in the uncanny valley territory, and it will be delayed for years due to resistance in the industry and legal battles over personality rights. One solution would be to create a completely fresh digital actor or base it on a low-paid unknown actor who will sign away their rights. But you can already just hire unknowns if you want to be cheap, and some films will even cast unknowns in the lead roles.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 1, Informative) by Anonymous Coward on Wednesday May 30 2018, @05:32AM (3 children)
I don't even see the appeal. Real people are hideous, including the ones who are commonly said to be beautiful. Fap to anime characters instead. No need for "Deepfakes".
(Score: 2) by takyon on Wednesday May 30 2018, @05:42AM (1 child)
*Slow clap*
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 1, Touché) by Anonymous Coward on Wednesday May 30 2018, @10:21AM
Slow Clap?, Cue the sound of one hand clapping..
(Score: 0) by Anonymous Coward on Wednesday May 30 2018, @06:13AM
A local skinflick house used to have an "Anime" festival late Sunday nights.
The one time I attended I did indeed fap to a fappable anime.
It was only after I jizzed all over the floor that I realized no one else was committing onanism. While I still hope this wasn't the case, I think most of the audience consisted of honest-to-goodness anime fans.
(Score: 0) by Anonymous Coward on Wednesday May 30 2018, @09:31AM (1 child)
takyon without citiations, sigh.
https://en.wikipedia.org/wiki/Final_Fantasy:_The_Spirits_Within [wikipedia.org]
And of course https://en.wikipedia.org/wiki/Hatsune_Miku [wikipedia.org]
(Score: 3, Touché) by takyon on Wednesday May 30 2018, @10:23AM
Sigh. Those are some "fine" "citations" you have there.
I think we can agree that Hatsune Miku isn't designed to be ultra realistic, avoiding the uncanny valley by going anime-style kawaiiiiii instead. Even the voice sounds unnatural, although that is also part of the appeal (refer to all of the auto-tuned music made in the last couple of decades).
A more timely example would be the virtual Peter Cushing and de-aged Carrie Fisher in Rogue One [wikipedia.org], a 2016 film:
You can also look at the Soylent reaction to "Siren":
Epic Games (Developer of the Unreal Engine) Shows Off "Siren" Demo [soylentnews.org]
The uncanny valley is deep and it will take some time to crawl out of it. Entirely computer-generated photorealistic actors/actresses are not routinely used. Motion capture in animation (rather than live action or faux live action) doesn't count because the characters are cartoony and unrealistic, or still fail the uncanny valley test [wikipedia.org]. Gollum/Sméagol as portrayed by Andy Serkis also doesn't count because the character design is exaggerated, fantastical, and meant to be discomforting.
The bottom line is that successful virtual acting means that audiences should not be able to distinguish whether or not a character is virtual. Ideally, motion capture would be optional. In fact, it should be possible to make a realistic film with no sets and cameras at all. These developments will have important implications for Hollywood as well as #deepfakenews.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by DavePolaschek on Wednesday May 30 2018, @12:35PM
It's M-m-m-max Headroom! https://en.wikipedia.org/wiki/Max_Headroom_(TV_series) [wikipedia.org]