Carlo Masone

Carlo Masone is a Fixed-term Researcher at the Department of Control and Computer Science (DAUIN), Politecnico di Torino, affiliated with the College of Electronic, Telecommunications, and Physics Engineering. He is also an invited member of the College of Management and Production Engineering and the College of Mathematical Engineering, reflecting his interdisciplinary contributions. His research focuses on artificial intelligence, machine learning, computer vision, and robotics, with significant work in visual place recognition, federated learning, and geospatial AI.

His research interests span Artificial Intelligence, Machine Learning, Computer Vision, Robotics, Visual Place Recognition, Federated Learning, Open-Set Segmentation, and Geospatial AI. He has developed methods for uncertainty quantification in ML models for autonomous vehicles and contributed to digital heritage through the MAPP project.

The recent publications highlight a strong trend in visual place recognition (e.g., EigenPlaces, MeshVPR), geolocalization of astronaut photography (Earthloc, EarthMatch), and robust segmentation (Mask2Anomaly). There is a clear emphasis on leveraging pre-trained features, federated learning for privacy-preserving recognition, and topological methods in vision. His work bridges robotics, AI, and real-world applications in autonomous systems and cultural heritage.

Scientific Awards:

• Outstanding Reviewer at CVPR (2022, 2023, 2024)
• IROS JTCF Novel Technology Paper Award (2016)
• Best Paper Award Finalist at IEEE ICIA (2016)
• ELLIS Fellow (2024–)

He advises PhD students in Artificial Intelligence and Computer and Systems Engineering and leads commercial research projects such as Quantifying Uncertainties in ML Models for Driver-Assisted and Autonomous Vehicles and MAPP - Phygital and Participatory Alpine Museums. He has previously collaborated with the Max Planck Institute for Biological Cybernetics (2014–2017).

He teaches courses including Machine Learning for Mathematical Engineering, Robot Learning, and GeoAI, and contributes to programming and data science curricula. He is also a co-inventor on a patent for automated neural network design.