A group of researchers from Tel Aviv University, in collaboration with representatives of the social network Facebook, has developed an algorithm that allows you to "revive" photos and portraits of people, transferring them to facial expressions from the reference video.

As N + 1 reminds, at present there are various methods that allow to collect data on the movements and facial expressions of a person, translate them into digital form, and then transfer them to a three-dimensional model of another person or character. Usually, such data are collected by means of special sensors or a set of markers for cameras fixed on the human body. This is the approach used to create computer graphics in films. Another group of methods makes it possible to do without additional equipment, but in this case, a large array of initial data is required.

The technology proposed by the Israeli researchers, allows you to "revive" one photo of a person or even a portrait using only one video with another person who demonstrates the desired facial expression, hitech.newsru.com reports. In the process, the algorithm transfers the person's facial expressions from the video to the photo or image, and to make the final result more realistic, the program also complements the animated image with details, including wrinkles that occur when the facial expression changes, or teeth that are exposed during a smile.

In the past year, the team of researchers from the Stanford University, University of Informatics, University of Plants, and the University of Erlangen-Nuremberg, have shown the technology to be transformed into a human-to-be video communication with a different human.