University of Bonn
2025 - Present
PhD in Computer Science, advised by Florian Bernard
Tobias Weißberg
PhD researcher at the University of Bonn, advised by Florian Bernard. I study structure and symmetry in geometry, computer vision, and neural networks.
Education
University of Bonn
- Graduated with distinction
2022 - 2024
Master's Degree in Computer Science
University of Bonn
2018 - 2022
Bachelor's Degree in Mathematics
Research Areas
Publications
Beyond Bit Matching: Orthogonal Watermarks for Collusion-Resistant Image Fingerprinting
Zhakshylyk Nurlanov, Tobias Weißberg, Florian Bernard
NeurIPS, 2026
Beyond Bit Matching: Orthogonal Watermarks for Collusion-Resistant Image Fingerprinting
Zhakshylyk Nurlanov, Tobias Weißberg, Florian Bernard
NeurIPS, 2026
Towards Stellarator Geometry Optimisation for Nuclear FusionTowards Stellarator Geometry Optimisation for Nuclear Fusion
Tobias Weißberg, Moritz Heep, Zorah Lähner, Florian Bernard
ECCV Workshop on Geometric Intelligence: From Vision to Scientific Discovery, September 2026 · Spotlight
Towards Stellarator Geometry Optimisation for Nuclear FusionTowards Stellarator Geometry Optimisation for Nuclear Fusion
Tobias Weißberg, Moritz Heep, Zorah Lähner, Florian Bernard
ECCV Workshop on Geometric Intelligence: From Vision to Scientific Discovery, September 2026 · Spotlight
We improve stellarator geometries for nuclear fusion through local shape refinement and a compact, two-dimensional design space.
Unsupervised Pixel-Level Semantic Left-Right Understanding of In-the-Wild ImagesUnsupervised Pixel-Level Semantic Left-Right Understanding of In-the-Wild Images
Weikang Wang, Tobias Weißberg, Florian Bernard
European Conference on Computer Vision (ECCV), September 2026
Unsupervised Pixel-Level Semantic Left-Right Understanding of In-the-Wild ImagesUnsupervised Pixel-Level Semantic Left-Right Understanding of In-the-Wild Images
Weikang Wang, Tobias Weißberg, Florian Bernard
European Conference on Computer Vision (ECCV), September 2026
We combine 3D shapes and images to learn which pixels belong to an object’s left or right side, without labelled training examples.
Symmetry Informative and Agnostic Feature Disentanglement for 3D ShapesSymmetry Informative and Agnostic Feature Disentanglement for 3D Shapes
Tobias Weißberg, Weikang Wang, Paul Roetzer, Nafie El Amrani, Florian Bernard
International Conference on 3D Vision (3DV), March 2026
Symmetry Informative and Agnostic Feature Disentanglement for 3D ShapesSymmetry Informative and Agnostic Feature Disentanglement for 3D Shapes
Tobias Weißberg, Weikang Wang, Paul Roetzer, Nafie El Amrani, Florian Bernard
International Conference on 3D Vision (3DV), March 2026
We disentangle 3D shape features into separate symmetry-informative and symmetry-agnostic components, improving symmetry detection and shape matching.
χ: Symmetry Understanding of 3D Shapes via Chirality Disentanglementχ: Symmetry Understanding of 3D Shapes via Chirality Disentanglement
Weikang Wang †, Tobias Weißberg †, Nafie El Amrani, Florian Bernard
IEEE/CVF International Conference on Computer Vision (ICCV), Oct 2025
χ: Symmetry Understanding of 3D Shapes via Chirality Disentanglementχ: Symmetry Understanding of 3D Shapes via Chirality Disentanglement
Weikang Wang †, Tobias Weißberg †, Nafie El Amrani, Florian Bernard
IEEE/CVF International Conference on Computer Vision (ICCV), Oct 2025
We extract left-right information from image foundation models to resolve symmetry ambiguities in 3D shapes.
Beyond Complete Shapes: A Benchmark for Quantitative Evaluation of 3D Shape Matching AlgorithmsBeyond Complete Shapes: A Benchmark for Quantitative Evaluation of 3D Shape Matching Algorithms
Viktoria Ehm †, Nafie El Amrani †, Yizheng Xie, Lennart Bastian, Maolin Gao, Weikang Wang, Lu Sang, Dongliang Cao, Tobias Weißberg, Zorah Lähner, Daniel Cremers, Florian Bernard
Computer Graphics Forum, Jun 2025
Beyond Complete Shapes: A Benchmark for Quantitative Evaluation of 3D Shape Matching AlgorithmsBeyond Complete Shapes: A Benchmark for Quantitative Evaluation of 3D Shape Matching Algorithms
Viktoria Ehm †, Nafie El Amrani †, Yizheng Xie, Lennart Bastian, Maolin Gao, Weikang Wang, Lu Sang, Dongliang Cao, Tobias Weißberg, Zorah Lähner, Daniel Cremers, Florian Bernard
Computer Graphics Forum, Jun 2025
We introduce a benchmark and dataset-generation framework for evaluating how well algorithms match complete and partial 3D shapes.
† denotes equal contribution.
Awards and Achievements
Prize for Outstanding Thesis, 2024Prize for Outstanding Thesis, 2024
Bonn Computer Science Society (e.V.) · June 2025
Prize for Outstanding Thesis, 2024Prize for Outstanding Thesis, 2024
Bonn Computer Science Society (e.V.) · June 2025
Awarded for my master’s thesis 'Learning Semantic Understanding of Human-Made Sketches', which advanced methods for machine learning–based interpretation of hand-drawn sketches.