Incoming M.Sc. Mathematics in Science and Engineering, TUM · 3D and 4D computer vision
I build computer vision systems that reconstruct moving 3D scenes and estimate when their predictions can be trusted.
My work covers 3D and 4D reconstruction, egocentric vision, multiple-view geometry, Gaussian Splatting, and uncertainty estimation.
Open to computer vision research and engineering roles in Munich, Zurich, and Paris.
Current research ongoing Dec 2025 to present
Reconstructing dynamic and egocentric scenes as their structure changes, and estimating when the result can be trusted.
Computer Vision Researcher, Bocconi University.
SE(3) Uncertainty for VGGT completed 2026
Extends VGGT with a separate branch that predicts a full 6x6 camera-pose covariance on the SE(3) manifold, while the backbone and the pose predictor stay frozen. The uncertainty ranks frames by their pose error and matches the structure of the covariance that bundle adjustment computes.
Sole author, BSc thesis. Supervised by Prof. Alessandro Pigati, Bocconi University.
Prosthetic Arm Vision public Sep 2024 – Jun 2025
A live pipeline combining object detection, monocular depth, segmentation, hand tracking, and Kalman-filtered 3D state. It feeds a model that judges whether a grasp is feasible from the object's 3D geometry and the user's hand pose, for a prosthetic arm controlled by EEG.
Computer Vision Lead and Project Lead, BAINSA and Politecnico di Milano.
October 2026 – September 2028
Technical University of Munich
M.Sc. Mathematics in Science and Engineering. Focus on robotics perception, multiple-view geometry, and 3D learning.
September 2023 – July 2026
B.Sc. Mathematics and Computing Sciences for Artificial Intelligence. Final grade 110/110 cum laude.
September 2025 – January 2026
Exchange semester. Grade point average 4.3/4.0, A+.
Kornia June 2026 – present
Vision Dental June 2025 – September 2025
Politecnico di Milano & Bocconi AI and Neuroscience Association September 2024 – June 2025
Accenture June 2025 – September 2025
My work combines mathematical geometry with practical vision systems, spanning 3D reconstruction, pose uncertainty, medical imaging, open-source computer vision, and real-time perception.
Graduated from Bocconi University in Mathematics and Computing Sciences for AI with 110/110 cum laude, with a thesis on uncertainty estimation for 3D foundation models. Thesis page
Two pull requests merged into Kornia: stereo disparity metrics, and TorchScript fixes across the geometry module.
Completed an exchange semester at École Polytechnique.