Pornhub began to remove fake celebrity videos generated by the neural network.

FakeApp program

On January 12, 2018, a Reddit user named Deepfakes released FakeApp , an application for replacing faces in videos. Since that time, several new versions have been released (the current version of FakeApp 2.1 was released on February 4) - and the program has become terribly popular. The number of downloads has long exceeded 100,000. Especially she liked the strawberry lovers, who have long wanted to see their favorite actress or some other chaste celebrity in an interesting position.

Fake celebrity videos were so widespread that popular video hosting had to take action. The first ban on fake videos was introduced by the famous site Pornhub - the global porn giant with an attendance of 75 million people a day.

“Users started tagging such content, and we remove it as soon as there are complaints,” said Pornhub Vice President Corey Price in a comment on Mashable . “We urge everyone who has this problem to visit our content removal page to officially submit a request.”
')
Later, representatives of Pornhub clarified their position. In the commentary for the Motherboard, they said they regarded it as a form of porn without the consent of the actress. And on the site, all types of videos were always blocked, when the participant clearly does not agree to what is happening. Now, this kind of content includes AI generated videos with strangers.

Although Pornhub has formally banned porn with celebrities, but in reality, dozens of such videos are uploaded to the site daily. And soon their quality will increase so much that it will become very difficult to distinguish: this is a real video with a celebrity or a fake one.

Deepfakes first showed the results of its program in December 2017. Then he has not yet posted it in open access. The author said that the imposition of faces is carried out by a system based on a neural network, which is developed using various free libraries and frameworks. He basically made the program using the Keras library on the TensorFlow backend . For learning neural networks and compiling celebrities' faces, it’s enough to search Google for images, stock photos and videos from YouTube.

The structure of the neural network is similar to that described in the scientific work of researchers from Nvidia . Their algorithm can, for example, in real time change the weather to video.


, . . /r/deepfakes/ .

, CPU . CUDA 8.0. , .

, - , . , . ffmpeg -i scene.mp4 -vf fps=[FPS OF VIDEO]. «» — , .

. .

. , (Grace Evangeline) , — . , . , .

, , — . , . , . -.

, , , . , , .

FakeApp 2.1 (), 1,6