M.Sc. Sebastian Weiß


Foto von Sebastian Weiß

M.Sc. Sebastian Weiß

Technische Universität München

Institut für Informatik

Postadresse

Postal:
Boltzmannstr. 3
85748 Garching b. München

Dienstort

Informatik 15 - Lehrstuhl für Grafik und Visualisierung (Prof. Westermann)

Work:
Boltzmannstr. 3(5613)/II
85748 Garching b. München

Publications

I am a PhD student in the group of Rüdiger Westermann. My research focuses on the acceleration of 3D isosurface and volume rendering. For that purpose, I investigate approximation, compression and deep learning techniques for their application in computer visualization. The goal is to be able to explore large datasets in real-time which would otherwise be difficult with traditional methods.

Differentiable Direct Volume Rendering
Sebastian Weiss, Rüdiger Westermann
to appear in IEEE VIS 2021
[PDF] [Preprint] [BIB] [Code

Analytic Ray Splitting for Controlled Precision DVR
Sebastian Weiss, Rüdiger Westermann
Eurographics Conference on Visualization (EuroVis) 2021, Best Short Paper
[PDF] [Preprint] [BIB] [Video] [Code] [Extended Version]

Deep Learning-Based Upscaling for In Situ Volume Visualization
Sebastian Weiss, Jun Han, Chaoli Wang, Rüdiger Westermann
In Hank Childs, Janine Bennett, and Christoph Garth (Eds.) In Situ Visualization for Computational Science, Springer, 2021, In Press
[PDF] [BIB]

Learning Adaptive Sampling and Reconstruction for Volume Visualization
S. Weiss, M. Işık, J. Thies, R. Westermann
TVCG 2020
[PDF] [Preprint] [BIB] [Source Code]

Volumetric Isosurface Rendering with Deep Learning-Based Super-Resolution
Sebastian Weiß, Mengyu Chu, Nils Thuerey, Rüdiger Westermann
TVCG 2019
[PDF][Preprint] [BIB] [Source Code] [Video]

Correspondence-Free Material Reconstruction using Sparse Surface Constraints
Sebastian Weiß, Robert Maier, Daniel Cremers, Rüdiger Westermann, Nils Thuerey
CVPR 2020
[PDF] [ArXiv Preprint ] [BIB] [Source Code] [Video]

Triplanar Displacement Mapping for Terrain Rendering
Sebastian Weiß, Florian Bayer, Rüdiger Westermann
Eurographics 2020
[Preprint] [BIB] [Source Code]

Visual Exploration of Circulation Rolls in Convective Heat Flows
A. Frasson, M. Ender, S. Weiss, M. Kanzler, A. Pandey, J. Schumacher, R. Westermann
IEEE Pacific Visualization 2019
[PDF ], [BIB]

Presentations / Tech Reports

Name Resources
PyTorch - CUDA - C++ Interop
(presentation held in January 2020)
[PDF] [Sources]
Inverse Volume Rendering + Introduction to Automatic Differentiation
(presentation held in November 2020)
[PDF]
Volume Rendering Integral: Derivation of the Front-to-back Algorithm
(tech report, December 2020)
[PDF]

Teaching

Bachelor and Master Thesis Topics

  • Numerics / Simulation on the GPU with CUDA
  • 3D visualization, especially Direct Volume Rendering
  • 3D graphics, especially terrain rendering
  • Deep learning in visualization
  • Combining music and computer science

If you are interested in one of the above topics, please contact me.

Supervised Theses

Topic Student Type of Thesis
2019    
Interpolation of Sprite Animation Using Deep Learning Daniel Drescher Bachelor
Deep Learning for Techno Music Manuel Geiger Bachelor
Triplanar Displacement Mapping for Terrain Rendering Florian Bayer Master
2020    
Multiple Precision Arithmetic on the GPU Dominic Ebner Bachelor
Perceptual Level-of-Detail for Physically-based Rendering Moritz Kohr Master
Real-time image inpainting on the GPU Nikita Fetisov Master
Analytic Ray-Isosurface Intersections Philipp Eckart-Weißbäcker Bachelor
Animating Mouth Movements from Text and Speech Florian Hartl Bachelor
Real-time Neural View Synthesis for Volumetric Datasets Parika Goel Master