Well now, it seems like you can’t claim to be in the business of next-gen technology these days until you’ve wheeled out a giant floating man-face. And yes, that's obviously due to the incredible challenges it takes to simulate a real 'life' head. Nvidia has now shown their own real-time face technology at their GPU Technology Conference in California. Check out the very impressive video after the break.
Face Works uses a face and motion capture technology developed at the University of Southern California’s Institute of Creative Technology (ICT). The Light Stage technology is able to capture data to within a tenth of a millimetre using photographic techniques that capture the geometry of an actor’s face as well as the light transmission through human skin and the reflections that come from the oils too. As well as the still images the Light Stage then uses video capturing to grab around thirty different facial expressions from the actor involved. Once all that information is captured it’s then compressed down to a set of meshes that can be used to recreate and animate a digitised version of that face.
Nvidia have worked with the ICT and managed to get this compressed data down from the 32GB it is in its raw form to around 400MB. At that size it is now possible to do the animation in real time and that is exactly what Jen-Hsun Huang has going on behind him on stage in the video below.
The demo is running a GTX Titan pretty hard, demanding around 2TFLOPS of processing power, around half the possible performance of the ultra-enthusiast GPU. It’s perhaps telling that the Face works tech is being demoed using a bald man – matching it with some proper follicle rendering technology would likely bring even the GTX Titan to its knees. Hmmm ... TressFX anyone ?