Imagine someone stealing your Facebook profile photo and making a fake video of yours, fortunately, the technology has not yet come into the hands of bad guys.
But Samsung has certainly learned how to do that.
Deep-faked videos are clips in which the onlookers are saying or doing what they have never done, and for this purpose, enough data is needed to produce real-looking footage.
Now Samsung has developed a new Artificial Facial Intelligence System that any fake clip can create with just one photo.
This technology is currently being used by the company for fun, such as giving life to classical paintings such as Mona Lisa and that is only possible with a single picture.
An artificial facial intelligence leap from Samsung-based Russia has developed the technology that was introduced this week.
Deep Fake software makes a human move and speaks to the extent of satisfying a viewer using machine learning technology, even though such changes have been made for decades from computer to video, making it easy to clip from the Deep Fake system. In fact, it is difficult to detect that they are fake.
While this technique is commonly used for online memes, Deep Fake technology is also dangerous because it can be used to indoctrinate a person's face into a pornographic film and is used in Europe or the United States.
In an article published by Samsung's AI Lab, it is described as a real neural talking head, of which the Talking Heads is a video created by the system, just like the video boxes appearing on TV News. , While neural refers to neural networks, a type of machine learning that attempts to imitate the human brain.
The lab has also released a video based on this technology and says it is not complete yet because only one or a few images have the details to make such a deep-faked video.