Can not dance? Artificial Intelligence helps you

An algorithm creates a virtual skeleton that maps the poses, while two other algorithms work to create a realistic image while dancing, explains Motherboard.

Researchers at the University of Berkeley have developed a deep learning system that picks up the dance steps of a source video and applies them to people with few skills for such moves.

An algorithm creates a virtual skeleton that maps the poses, while two other algorithms work to create a realistic image while dancing, explains Motherboard.

To obtain reference material, it is necessary for the person to dance (or move freely through space) alone at the beginning.

The system is still at an early stage, as can be seen in the video where the technology is sampled. It is that in these images, it is possible to see that the movement is not yet fully fluid and, on some occasions, one even notices the lack of a forearm or other parts of the body. In addition, replicate dances are relatively simple, as entering into worlds like breakdance raises other challenges, as Engadget points out.

It remains to be seen if this technology will be able to withstand a fate similar to that of deepfakes, in which something that allowed the actors to change faces in movies was used in an abusive way to create false pornography with celebrities.