Tech giant Samsung Electronics has developed a new artificial intelligence technology that can turn still images into a video which shows various facial expressions and movements of a person in the picture. The new technique focuses on using just a few photographs or even just one still image.
There are other technologies that can do the same but the others require a large number of images of the person that you want to animate. But Samsung's technology presented in a paper called Few-Shot Adversarial Learning of Realistic Neural Talking Head Models can turn a single still image into a video. The technology uses facial features from a publicly available database of more than 7000 images of celebrities and a huge number of videos of people talking to the camera.
A video explaining Samsung's AI tech is also available online and as a demonstration in the video, the ever popular Mona Lisa from the hands of Leonardo da Vinci can be seen giving facial expressions. The Mona Lisa is animated using three different source videos that can produce very different results both in facial structure and behaviour. The tech video has gone viral on social media.
Here is the viral video and the explanation of the new tech,
Few-Shot Adversarial Learning of Realistic Neural Talking Head Models