Researchers at the Samsung’s AI Center in Moscow have managed to use Deepfake technology to create “living portraits” from as few as just a single photograph. The paper, titled “Few-Shot Adversarial Learning of Realistic Neural Talking Head Models,” was published on the preprint server arXiv on Monday. Simply put, the AI can be trained using just one image to create a realistic-looking, animated portrait, and by adding a few more shots (up to 32 photographs) the realism improves. Read more for a video and additional information.
“Because they only need one source image, the researchers were able to animate paintings and famous portraits, with eerie results. Fyodor Dostoevsky—who died well before motion picture cameras became commercially available—moves and talks in black and white. The Mona Lisa silently moves her mouth and eyes, a slight smile on her face. Salvador Dali rants on, mustache twitching,” reports Vice.