Senior Scientist and Lecturer, ETH Zürich
Computer Vision · 3D/4D Reconstruction · Spatial Intelligence
I study how machines build internal models of the physical world, from 3D geometry and 4D dynamics to structured representations for spatial reasoning.
I am a senior scientist and lecturer at ETH Zürich, working across computer vision, computer graphics, and machine learning. My research focuses on representations of real-world 3D and 4D phenomena: reconstructing geometry, modelling motion, rendering photorealistic scenes, and extracting structure from visual observations.
I am increasingly focused on spatial intelligence: the bridge between geometric reconstruction and higher-level reasoning. My goal is to build models that form and update internal maps to answer three questions: what is where, what changes, and how a machine can interact with its environment.
I completed my PhD in Michael Black's Perceiving Systems department at the Max Planck Institute for Intelligent Systems, supervised by Peter Gehler and Sebastian Nowozin. Before that, I spent seven years (2008–2015) building machine learning systems at scale for malware detection at Kaspersky Lab.
Methods for reconstructing and rendering photorealistic static and dynamic scenes from sparse or imperfect observations.
Efficient representations for geometry, appearance, motion, and time-varying real-world structure.
Digital representations of humans, biological systems, and surgical scenes: controllable, temporally coherent, and grounded in real observations.
Understanding what kinds of internal maps vision models need to reason about space, motion, and change.
At ETH Zürich, I co-teach Computer Vision (with Marc Pollefeys and Siyu Tang) and Digital Humans (with Siyu Tang).
I supervise student projects in computer vision, 3D reconstruction, neural rendering, and spatial intelligence, with particular interest in work that connects rigorous geometric modelling with foundation models.
If you are a student interested in working on these topics, please reach out. The most useful messages are short and specific: what you want to work on, why it interests you, and what you have already tried.