Home Pub.s Proj.s   CV  


Our new group website is online now at [ge.in.tum.de]. Please check it out for our latest resarch and updates!



In particular, you can find info about our CNNs for physics simulations, and about our fluid solver [mantaflow] there. There are also tutorials on how to use it for training deep net's / CNNs with [tensorflow].

My "old" homepage:

I am Assistant Professor at [TU Munich]. Before that I was at [ETH Zurich], and [ScanlineVFX]. My publications can be found [here], while this is my [CV], and these are some additional publication stats: [1], [2], [3]. You can contact me at: "nils (dot) thuerey (at) tum.de".



Mission statement:
The focus of my research is to develop numerical methods for physics simulations in computer graphics. A particular emphasis lies on simulating fluid flow with deep learning algorithms. These methods have significant potential for efficient and robust simulations in the computer graphics and digital games context, but I am also very interested in applications from the fields of mechanical engineering and medicine.



Updates:
- 2017-11-21: Shutting down, please check out my TUM hompeage instead
- 2017-08-23: our Turbulent Fluids course has moved to the GE homepage
- 2017-05-04: Our siggraph papers are also finally available
- 2017-04-27: Our real-time liquids paper with OF and neural networks is online now
- 2017-03-14: Our PD optimization paper is available now
- 2017-02-01: Data-driven fluids are online, with source code
- 2014-04-10: Update for our Turbulent Fluids course
to a new website




ERC StG realFlow