Zum Inhalt springen
Menü
Technische Universität München
Technische Universität München

Master-Seminar - Deep Learning in Physics (IN2107, IN0014)

LecturerMarie Lena Eckert, Nils Thuerey
StudiesMaster Informatics
Time, Place

Seminarraum MI  02.13.010

Wednesday, Nov. 14., 2018 12:00-16:00
Wednesday, Nov. 21., 2018 12:00-16:00
Wednesday, Nov. 28., 2018 12:00-16:00
[Wednesday, Dec. 05., 2018 12:00-16:00] 

Begin15 October 2018

Content

In this course, students will autonomously investigate recent research about machine learning techniques in the physical simulation area. Independent investigation for further reading, critical analysis, and evaluation of the topic are required.

This seminar takes place in the specified 3 or 4 time slots. In each time slot, 4 to 5 students will present their paper. Everyone needs to attend all time slots. If you partially can't participate one time slot, please get in contact with us in advance.

Requirements

  • The participants have to present their topics in a talk (in English), which should last 30 minutes. Don't put too many technical details into the talk, make sure the audience gets the paper's main idea. Be prepared to answer questions regarding the technical details, you could prepare backup slides for that.
  • Plagiarism is important; please do not simply copy the original authors' slides. You can certainly refer to them.
  • The semi-final slides (PDF) should be sent one week before the talk; otherwise, the talk will be canceled. We strongly encourage you to finalize the semi-final version as far as possible. We will take a look at the version and give feedback. You can revise your slides until your presentation.
  • A short report (approximately 3-4 pages in the ACM SIGGRAPH TOG format (acmtog)) should be prepared and sent within two weeks after the talk. When you send the report, please send the final slides (PDF) as well.
  • Guideline: You can begin with writing a summary of the work you present as a start point; but, it would be better if you focus more on your own research rather than just finishing with the summary of the paper. We, including you, are not interested in revisiting the work done before; it is more meaningful if you make an effort to put your own reasoning about the work, such as pros and cons, limitation, possible future work, your own ideas for the issues, etc.

Papers

Date Presenter Paper
Nov. 14.Stephan Pirner2016, Tompson et al., Accelerating Eulerian Fluid Simulation With Convolutional Networks, arXiv.org
Nov. 21.David Wagner2016, Li et al., To Fall Or Not To Fall: A Visual Approach to Physical Stability Prediction, arXiv.org
Nov. 21.Moritz Kohr2017, Watters et al., Visual Interaction Networks, arXiv.org
Nov. 14.Abraham Duplaa2018, Schenck, SPNets: Differentiable Fluid Dynamics for Deep Neural Networks, arXiv.org
Nov. 28.Bakar Andguladze2018, Luo et al., DeepWarp: DNN-based Nonlinear Deformation, arXiv.org
Nov. 14.Georg Kohl2015, Ladicky et al., Data-driven Fluid Simulations using Regression Forests, ACM Trans. Graph.
Nov. 21.Konstantin Kraus2018, Kim et al., Deep Fluids: A Generative Network for Parameterized Fluid Simulations, arXiv.org
Nov. 28.Abdelkader Saad2016, Holden et al., A Deep Learning Framework for Character Motion Synthesis and Editing, ACM Trans. Graph.
Nov. 21.Nicolas Pinkau2018, Erhardt et al., Unsupervised Intuitive Physics from Visual Observations, arXiv.org
Nov. 28.Maximilian Schmidt2018, Baque et al., Geodesic Convolutional Shape Optimization, arXiv.org
--2018, Pathak et al., Model-Free Prediction of Large Spatiotemporally Chaotic Systems from Data: A Reservoir Computing Approach, Physical review letters
Nov. 14.Akshay Tumkur Renukaprasad2018, Beck et al., Deep Neural Networks for Data-Driven Turbulence Models, arXiv.org
Nov. 28.Jan Fahlbusch2018, Bailey et al., Fast and deep deformation approximations, ACM Trans. Graph.
Nov. 14.Yang Liu2018, Y. Xie et al., tempoGAN: A Temporally Coherent, Volumetric GAN for Super-resolution Fluid Flow, ACM Trans. Graph.
Nov. 21.Daniil Zauzolkov2018, Wiewel et al., Latent-space Physics: Towards Learning the Temporal Evolution of Fluid Flow, arXiv.org
--2016, C. Yang et al., Data-driven projection method in fluid simulation, Computer Animation and Virtual Worlds

You can access the papers through TUM library's eAccess.

 

References