Hefei University of Technology has developed software that uses facial recognition to analyze movies to produce a comic book based on the movie.
Developed at the Hefei University of Technology, in China, the clever tech will first of all analyse a film, identify the characters via facial recognition, and then start taking screen-grabs of the action.
Once this is complete, it can shuffle and order the screen-grabs to tell a coherent story through the medium of the comic panel format.
And then finally it can add speech bubbles, and render the finished images in the chosen style, for instance as a cartoon-or as anime.
The Daily Mail notes that it’s only 85% accurate.