NVIDIA engineers have developed an AI-powered neural network capable of generating ultra realistic virtual characters in real-time. Technically speaking, a neural network-based simulation super-resolution framework that can efficiently and realistically enhance a facial performance produced by a low-cost, real-time physics-based simulation.
The team used facial animations as an exemplar of such a simulation domain, where creating this agreement is achieved by dialing in the same muscle actuation controls and skeletal pose in the two simulators. The neural network super-resolution framework then generalizes from this training set to unseen expressions, while also compensating for modeling discrepancies between the two simulations due to limited resolution or cost-cutting approximations in the real-time variant.
- ELITE GAMING & CREATION ANYWHERE: The Intel Core i7-13650HX processor goes beyond performance to let your PC do even more at once. With a...
- BEYOND FAST: Feel the power emanate from within the neon-hued shell of the Helios Neo 16. The NVIDIA GeForce RTX 4060 Laptop GPU is beyond fast for...
- STUNNING VISUALS: Lock your sights on this display. At 16 inches with a 16:10 aspect ratio,165Hz refresh rate, and 3ms response time the WUXGA LED...
Our proposed neural network super-resolution framework generalizes from this training set to unseen expressions, compensates for modeling discrepancies between the two simulations due to limited resolution or cost-cutting approximations in the real-time variant, and does not require any semantic descriptors or parameters to be provided as input, other than the result of the real-time simulation,” said the researchers.
