Here you can find some of the research projects in which I have been working during the last years.

Towards Depth-Guided Self-Supervised World Models (Feb. 2025 - Jul. 2025) Developed a self-supervised learning pipeline that distills geometric information via depth signals to enhance visual representations, which is crucial for future scene prediction in autonomous driving world models. This was part of the 3D Vision course taught by the Computer Vision and Geometry Group (CVG) at ETH Zürich. (Read More)

Gaussian Belief Propagation for Continuous-Time SLAM (Feb. 2025 - Jul. 2025) Systematically analyzed and solved numerical instabilities in the Hyperion solver, a distributed, real-time framework for Continuous-Time SLAM, by implementing and evaluating several regularization techniques. Semester project developed at the Vision for Robotics Lab (V4RL) at ETH Zürich. (Read More)

Event-based Egomotion Estimation (Feb. 2024 - Jul. 2024) Designed and implemented a geometric model-based system for estimating the egomotion of an event camera, leveraging contrast maximization for depth and plane-fitting for normal flow. This was the project developed during my bachelor’s thesis at IRI-CSIC (Institut de Robòtica i Informàtica Industrial). (Read More)