Research / Zurich
Norman Müller
Member of the Technical Staff at Microsoft AI
AI research, from new models to working systems.
Microsoft AI Zurich
About me
I’m a researcher working on reinforcement learning for agentic models at Microsoft AI in Zurich. My expertise also spans controllable video generation and 3D generative AI. Previously, I led team research on 3D GenAI at Meta Reality Labs Zurich and helped bring photorealistic reconstruction to Meta Horizon Hyperscape.
I completed my PhD at the Technical University of Munich, advised by Prof. Matthias Niessner. My dissertation is Generative Models on 3D Representations. Before that, I earned Bachelor’s and Master’s degrees in both Computer Science and Mathematics at RWTH Aachen University, graduating at the top of my class.
My background also includes robotics at BMW and research internships with Peter Kontschieder at Meta. I build tools for research, including DVIS for browser-based 3D visualization.
News & updates
All updates- LatestMAI-Thinking-1 is out! I contributed to its agentic coding capabilities, achieving state-of-the-art performance in its weight class.
- Our paper LuxRemix: Lighting Decomposition and Remixing for Indoor Scenes was accepted to CVPR 2026!
- I started at Microsoft AI as a Member of the Technical Staff!
- MapAnything accepted to 3DV 2026!
-
Selected publications & releases
-
MAI-Thinking-1
Contributed to its agentic coding capabilities, achieving state-of-the-art performance in its weight classCitation
@misc{microsoft2026maithinking1, title = {{MAI-Thinking-1}}, author = {{Microsoft AI}}, year = {2026}, note = {Contributed to its agentic coding capabilities, achieving state-of-the-art performance in its weight class} } -
LuxRemix: Lighting Decomposition and Remixing for Indoor Scenes
Edit individual light sources in a captured scene, consistently across viewpoints.
Citation
@inproceedings{liang2026luxremix, title = {LuxRemix: Lighting Decomposition and Remixing for Indoor Scenes}, author = {Liang, Ruofan and Müller, Norman and Weber, Ethan and Zauss, Duncan and Vijaykumar, Nandita and Kontschieder, Peter and Richardt, Christian}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, year = {2026}, url = {https://luxremix.github.io/}, } -
MapAnything: Universal Feed-Forward Metric 3D Reconstruction
Recover metric 3D geometry from images and optional camera or depth inputs.
Citation
@inproceedings{keetha2026mapanything, title = {{MapAnything}: Universal Feed-Forward Metric {3D} Reconstruction}, author = {Keetha, Nikhil and M\"{u}ller, Norman and Sch\"{o}nberger, Johannes and Porzi, Lorenzo and Zhang, Yuchen and Fischer, Tobias and Knapitsch, Arno and Zauss, Duncan and Weber, Ethan and Antunes, Nelson and Luiten, Jonathon and Lopez-Antequera, Manuel and Bul\`{o}, Samuel Rota and Richardt, Christian and Ramanan, Deva and Scherer, Sebastian and Kontschieder, Peter}, booktitle = {International Conference on 3D Vision (3DV)}, year = {2026}, organization = {IEEE}, url = {https://arxiv.org/abs/2509.13414}, } -
FlowR: Flowing from Sparse to Dense 3D Reconstructions
Citation
@article{fischer2025flowr, author = {Fischer, Tobias and Bul{\`o}, Samuel Rota and Yang, Yung-Hsu and Keetha, Nikhil Varma and Porzi, Lorenzo and M{\"u}ller, Norman and Schwarz, Katja and Luiten, Jonathon and Pollefeys, Marc and Kontschieder, Peter}, title = {{FlowR}: Flowing from Sparse to Dense 3D Reconstructions}, journal = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) [Highlight]}, year = {2025}, url = {https://arxiv.org/abs/2504.01647}, eprint = {2504.01647} } -
Easy3D: A Simple Yet Effective Method for 3D Interactive Segmentation
Citation
@article{simonelli2025easy3d, title = {Easy3D: A Simple Yet Effective Method for 3D Interactive Segmentation}, author = {Simonelli, Andrea and M{\"u}ller, Norman and Kontschieder, Peter}, journal = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) [Oral]}, year = {2025}, url = {https://arxiv.org/abs/2504.11024}, eprint = {2504.11024} } -
Generative Gaussian splatting: Generating 3D scenes with video diffusion priors
Citation
@article{schwarz2025generative, title = {Generative Gaussian splatting: Generating 3D scenes with video diffusion priors}, author = {Schwarz, Katja and M{\"u}ller, Norman and Kontschieder, Peter}, journal = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)}, year = {2025}, url = {https://arxiv.org/abs/2503.13272}, eprint = {2503.13272} } -
Coherent 3D Scene Diffusion From a Single RGB Image
Citation
@inproceedings{dahnert2024coherent, title = {Coherent 3D Scene Diffusion From a Single {RGB} Image}, author = {Dahnert, Manuel and Dai, Angela and M{\"u}ller, Norman and Nie{\ss}ner, Matthias}, booktitle = {The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year = {2024}, url = {https://openreview.net/forum?id=lckAdnVzsT}, } -
MultiDiff: Consistent Novel View Synthesis from a Single Image
Citation
@inproceedings{muller2024multidiff, title = {MultiDiff: Consistent Novel View Synthesis from a Single Image}, author = {M\"uller, Norman and Schwarz, Katja and R\"ossle, Barbara and Porzi, Lorenzo and Bul\`o, Samuel Rota and Nie{\ss}ner, Matthias and Kontschieder, Peter}, year = {2024}, month = jun, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)} } -
ViewDiff: 3D-Consistent Image Generation with Text-To-Image Models
Citation
@inproceedings{hollein2024viewdiff, title = {ViewDiff: 3D-Consistent Image Generation with Text-To-Image Models}, author = {H{\"o}llein, Lukas and Bo\v{z}i\v{c}, Alja\v{z} and M{\"u}ller, Norman and Novotny, David and Tseng, Hung-Yu and Richardt, Christian and Zollh{\"o}fer, Michael and Nie{\ss}ner, Matthias}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, year = {2024}, } -
Ganerf: Leveraging discriminators to optimize neural radiance fields
Citation
@article{roessle2023ganerf, title = {Ganerf: Leveraging discriminators to optimize neural radiance fields}, author = {Roessle, Barbara and M{\"u}ller, Norman and Porzi, Lorenzo and Bul{\`o}, Samuel Rota and Kontschieder, Peter and Nie{\ss}ner, Matthias}, journal = {ACM Transactions on Graphics (TOG)}, volume = {42}, number = {6}, pages = {1--14}, year = {2023}, publisher = {ACM New York, NY, USA} } -
DiffRF: Rendering-Guided 3D Radiance Field Diffusion
Generate 3D radiance fields directly with rendering-guided diffusion.
Citation
@inproceedings{muller2023diffrf, title = {DiffRF: Rendering-Guided 3D Radiance Field Diffusion}, author = {M{\"u}ller, Norman and Siddiqui, Yawar and Porzi, Lorenzo and Bul{\`o}, Samuel Rota and Kontschieder, Peter and Nie{\ss}ner, Matthias}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) [Spotlight]}, pages = {12568--12577}, year = {2023}, } -
Panoptic Lifting for 3D Scene Understanding with Neural Fields
Citation
@inproceedings{siddiqui2022panoptic, title = {Panoptic Lifting for 3D Scene Understanding with Neural Fields}, author = {Siddiqui, Yawar and Porzi, Lorenzo and Bul{\'o}, Samuel Rota and M{\"u}ller, Norman and Nie{\ss}ner, Matthias and Dai, Angela and Kontschieder, Peter}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) [Spotlight]}, pages = {12568--12577}, year = {2023}, } -
AutoRF: Learning 3D Object Radiance Fields From Single View Observations
Learn 3D object representations from single-view observations.
Citation
@inproceedings{Muller_2022_CVPR, title = {AutoRF: Learning 3D Object Radiance Fields From Single View Observations}, author = {M\"uller, Norman and Simonelli, Andrea and Porzi, Lorenzo and Bul\`o, Samuel Rota and Nie{\ss}ner, Matthias and Kontschieder, Peter}, year = {2022}, month = jun, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, pages = {3971--3980} } -
Seeing Behind Objects for 3D Multi-Object Tracking in RGB-D Sequences
Citation
@inproceedings{mueller2021completetracking, title = {Seeing Behind Objects for 3D Multi-Object Tracking in RGB-D Sequences}, author = {M{\"u}ller, Norman and Wong, Yu-Shiang and Mitra, Niloy J. and Dai, Angela and Nie{\ss}ner, Matthias}, year = {2021}, booktitle = {Proc. Computer Vision and Pattern Recognition (CVPR), IEEE} }