Samsung Develops Tech That Makes Still Portraits “Talk”

Last Updated on by

AI – what can’t it do?

Not much anymore and pretty soon it is going to be doing things that some of us never had the creativity to imagine before, such as this story.

  • Claim Your Free Camera Craft Cheat Sheet

Print it out and keep it for when you really need it - when you're out shooting!

arrow-circle-right
Image by Markus Spiske temporausch.com from Pexels

Have you ever wanted to bring the Mona Lisa to life? What about any other kind of painting or “still portrait?”

You might be able to do just that if Samsung’s recently patented technology takes off.

Basically what it does is that it places your facial movements overtop of a classical painting, making it come to “life” in the process.

Developed by the Samsung AI Center in Moscow in collaboration with the Skolkovo Institute of Science and Technology, the tech “generates 3D animated heads from a single still image. Unlike previously detailed AI systems capable of generating photo-realistic portraits, the new technology produces moving, talking heads” according to DPReview.

The program does this through adversarial networks that “fill in the details” that are missing. Tasked with developing this kind of software using only one picture as a reference, the team demonstrated that “…only a handful of photographs (as little as one) is needed to create a new model, whereas the model trained on 32 images achieves perfect realism and personalization score in our user study (for 224p static images).”





As of press, this is just the facial and facial animation, not the entire figure though animating still figures in images might be a future goal of this research.

We’ve talked a lot about AI technology in photography over the past year – and for good reason. From everything including software that can change the weather in a picture to tech that can create a composite landscape or portrait out of whole cloth, AI will become the next big thing in photography and it could be a huge challenge for some of us out there that hate even photoshopped or edited photographs.

You can watch a video demonstrating it by clicking here.

The question becomes: If AI can generate images and scenes of outstanding beauty or novelty, what incentive will the vast majority of us have to pack up our equipment and try to capture that in nature or out in the real world?

As always, we’d love to know your thoughts in the comments section below.

[DPReview]

About the author

    Kehl Bayern

    Kehl is our staff photography news writer and has over a decade of experience in online media and publishing and you can get to know him better here

    Leave a comment: