My research targets physical simulations for virtual worlds such as digital games, and movies. A particular focus of my work are simulations with deep learning techniques for effects such as detailed smoke and liquids. You can find some examples of SIGGRAPH papers that I worked on below. Neural networks are an exciting area for research, even more so when going beyond the established paths of vision and imaging problems.
If you’re interested in bachelor or master theses along these lines, please contact me. There are currently several interesting topics available in this area.
How to get in Touch
- eMail: nils.thuerey (at) tum.de
- Phone: +49 89 289 19484
- Fax: +49 89 289 19462
- Room: 02.13.061
Full CV: PDF.
Some statistics: Google-scholar, MS-academic, DBLP.
And here are some previews of selected works. You can check out the full list on our publications page. Among others, we’ve worked on for super-resolution fluids with a physics-based GAN:
And adversarial video super-resolution with temporal coherence using deep learning:
Our SIGGRAPH paper on flow descriptors from convolutional neural networks (numerical-viscosity-aware) for fluid flow:
Preceptual evaluations of physics simulation methods, independent of rendering styles:
For fun – here’s also an older animation I did with our fluid control algorithm and Blender: