News

Excited to give a talk at the BLISS Speaker Series on June 16, 2026 at 7PM. My talk will be titled "Beyond Patches: Learning Dense Visual Features". Slides are available here.

Happy to share that I will start a new position as Student Researcher at Google in Zurich in June 2026.

New pre-print: "AnyUp: Universal Feature Upsampling" is now available on arXiv! Super excited to share this work, where we propose a first-of-its-kind feature-agnostic upsampling architecture that can upsample features from any vision model at any resolution, without requiring any encoder-specific training. New state-of-the-art results on multiple downstream benchmarks, while being the first upsampler that naturally generalizes to different feature types at inference time. Accepted to ICLR 2026 as an oral presentation!

New pre-print: "Do It Yourself: Learning Semantic Correspondence from Pseudo-Labels" is now available on arXiv! We show that foundational features can be refined with an adapter that is trained with pseudo-labels, which are themselves zero-shot predictions using the same foundational features. We improve the quality of pseudo-labels through 3D-aware chaining with cycle-consistency and reject wrong pairs using a spherical prototype. New state-of-the-art results on SPair71k and scalable to larger datasets. Accepted to ICCV 2025!

Our pre-print "MEt3R: Measuring Multi-View Consistency in Generated Images" is now available on arXiv! In this work, we propose a DUSt3R-based method to measure multi-view consistency which can, e.g., be used to evaluate the 3D consistency of video diffusion models. Accepted to CVPR 2025!

Happy to report that my latest paper, "Gaussians-to-Life: Text-Driven Animation of 3D Gaussian Splatting Scenes", was accepted for publication at 3DV 2025. Thanks for a great collaboration to my co-authors, Michael Oechsle, Michael Niemeyer, and Federico Tombari!

I recently completed and defended my Master's thesis at TUM, marking the end of my Master's degree in Munich. 🎓

I 'll be joining the groups of Christian Theobalt, Bernt Schiele (both MPI), and Siyu Tang (ETH) for a PhD through the Max Planck ETH Center for Learning Systems (CLS) and ELLIS program later this year. Excited for what's to come! 🚀

Proud to announce that my paper Back to 3D: Few-Shot 3D Keypoint Detection with Back-Projected 2D Features just got accepted to CVPR 2024. Thanks to my co-authors, Maks Ovsjanikov and Peter Wonka, for the great collaboration.

Happy to announce that my master's thesis at TUM will be under the supervision of Federico Tombari and in collaboration with his team at Google Research, including Michael Niemeyer and Michael Oechsle. Looking forward to this new challenge!

I am glad to present my paper "Scale-Equivariant Deep Learning for 3D Data" at the first joint conference of all Konrad Zuse Schools of Excellence in AI. The poster that I am presenting in the common poster session can be found here.

The work on in-game goal prediction that I did together with Steffen Lang, Alexander Isenko and Daniel Link was presented at the 28. Annual Conference of the Commission on Soccer of the German Association for Sports Science (dvs). I am happy to announce that it was awarded with the young researcher award of the commission.

With the end of my internship in the group of Maks Ovsjanikov, my Master's studies in Paris come to an end. The studies in Paris were an enriching experience, and I am very grateful for the opportunity to work with so many great people. I am now looking forward to my last semester in Munich to finish my Master's degree there.

Happy to announce that I was selected for one of the prestigious Konrad Zuse School of Excellence in Learning and Intelligent Systems (ELIZA) research-oriented Master's scholarships that are awarded for outstanding talents in AI in Germany.

Our paper "Language Models for German Text Simplification: Overcoming Parallel Data Scarcity through Style-specific Pre-training" has been accepted to Findings of ACL 2023 and is now available on arXiv.

Our paper "Joint MR sequence optimization beats pure neural network approaches for spin-echo MRI super-resolution" is now available as a pre-print on arXiv. Check it out!

My first first-authored paper on "Scale-Equivariant Deep Learning for 3D Data" is now available as a pre-print on arXiv. Check it out!.