Photo Wake-up
posted June 6, 2019 #
This Photo Wake-Up project is blowing up my tiny little brain. Given a single photo input, a 3D model can be generated and animated. The above demo video is full of examples, all insanely impressive. They even go as far as to explain, broadly, what is happening behind the scenes.
Apparently tech like this has existed since at least 2007 but this newest take on the idea is the most robust and can even exist within a full AR space.
The examples they use are, surely, best case scenarios (i.e. no overlapping body parts, very crisp images, etc) but that doesn't make the concept any less bonkers. I've watched this four times at least.
Apparently tech like this has existed since at least 2007 but this newest take on the idea is the most robust and can even exist within a full AR space.
The examples they use are, surely, best case scenarios (i.e. no overlapping body parts, very crisp images, etc) but that doesn't make the concept any less bonkers. I've watched this four times at least.
