Fast Neural Network Emulation of Dynamical Systems for Computer Animation

Part of Advances in Neural Information Processing Systems 11 (NIPS 1998)

Bibtex Metadata Paper


Radek Grzeszczuk, Demetri Terzopoulos, Geoffrey E. Hinton


Computer animation through the numerical simulation of physics-based graphics models offers unsurpassed realism, but it can be computation(cid:173) ally demanding. This paper demonstrates the possibility of replacing the numerical simulation of nontrivial dynamic models with a dramatically more efficient "NeuroAnimator" that exploits neural networks. Neu(cid:173) roAnimators are automatically trained off-line to emulate physical dy(cid:173) namics through the observation of physics-based models in action. De(cid:173) pending on the model, its neural network emulator can yield physically realistic animation one or two orders of magnitude faster than conven(cid:173) tional numerical simulation. We demonstrate NeuroAnimators for a va(cid:173) riety of physics-based models.