Fast Neural Network Emulation of Dynamical Systems for Computer Animation

Abstract

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.

Cite

Text

Grzeszczuk et al. "Fast Neural Network Emulation of Dynamical Systems for Computer Animation." Neural Information Processing Systems, 1998.

Markdown

[Grzeszczuk et al. "Fast Neural Network Emulation of Dynamical Systems for Computer Animation." Neural Information Processing Systems, 1998.](https://mlanthology.org/neurips/1998/grzeszczuk1998neurips-fast/)

BibTeX

@inproceedings{grzeszczuk1998neurips-fast,
  title     = {{Fast Neural Network Emulation of Dynamical Systems for Computer Animation}},
  author    = {Grzeszczuk, Radek and Terzopoulos, Demetri and Hinton, Geoffrey E.},
  booktitle = {Neural Information Processing Systems},
  year      = {1998},
  pages     = {882-888},
  url       = {https://mlanthology.org/neurips/1998/grzeszczuk1998neurips-fast/}
}