Near-realtime Facial Animation by Deep 3D Simulation Super-Resolution
DescriptionWe present a neural network-based simulation super-resolution framework that enhances facial performance from a low-cost, real-time simulation to a detail level similar to high-resolution, offline simulators. By training on paired low- and high-resolution frames, our approach generalizes to unseen expressions and compensates for modeling discrepancies without needing semantic descriptors.
Event Type
Technical Papers
TimeTuesday, 3 December 20249:00am - 12:00pm JST
LocationHall C, C Block, Level 4
