General, Research, Technology

The Russian developer has created a new generation of the Deepfake algorithm. What can he do

Over the past few years, deep learning(Deep Learning) has gone a long way. Some systems of machine vision, voice perception and some others work so efficiently that they even sometimes surpass the capabilities of a person. But where there is a new promising technology, there will certainly be those who will try to use it for other purposes. This is how Deepfake - neural networks that can replace a faceperson in a photo or video. That is, any person will be able to sit in front of the camera, say something, and his face will be replaced by another person. But what if I told you that Deepfake can replace not only the face, but even hair, skin color and figure?

Still from a video taken with Deepfake 2.0

What the new Deepfake can do

According to Russian-speaking developer Alexey Chaplygin, he spent almost two years of work to create Deepfake new generation... His version of the algorithm can change almost everything: from skin and hair color to face and figure. In addition, the neural network is capable of changing a person in any photo, even if it was not initially loaded for processing by the algorithm. Here are examples of what it looks like.

How long has Nicolas Cage starred in Superman?

According to a recent report from MassachusettsInstitute of Technology, a device that allows the use of deepfakes could be "the perfect weapon for fake news creators who want to influence everything from stock prices to elections."

If you are interested in news of neural networks, subscribe to us in Yandex.Zen, so as not to miss new materials!

Why Deepfake is dangerous

You can talk as much as you like that suchtechnology is needed, it's very cool and you don't have to slander. You can even go to the extreme and begin to say that this is the position of a fierce oldfag who is simply afraid of everything new, but there really are more dangers than benefits. Just imagine if someone uses such a neural network and "records" a video featuring Donald Trump, in which he declares a trade war on China? Until the presidential administration refutes this, the US stock market will simply collapse. And behind it are the dollar and the reserves of many countries.

How recognize the deepfake? If you watch the video at the beginning of the article, thenyou will understand that while a fake can be literally with the naked eye. The picture is similar, but rather rough. In addition, she sometimes has some blending problems, especially around the edges of the face. But given that work on the algorithm is ongoing, there is a risk that soon it will be much more difficult to distinguish a fake video created by a neural network from a real one.

While deepfakes are easy to distinguish from the original, but soon it will be more difficult

You can imagine the opposite situation, when a real person says something, and then assures everyone that he was framed.

How to be in this situation is also not very clear. This will bring such confusion to the news feeds that it will simply not be possible to double-check it in another source. As a result, it will become generally unclear what is true and what is false in this world.

An algorithm that alreadycreate YouTube and Facebook. In the case of famous people, such a problem will be solved at the level of the video service, since the same YouTube knows how the conventional Donald Trump moves. But when it comes to a lesser known person, it gets trickier.