Deep Learning for Point Clouds Classification in the Ducal Palace at Urbino

A groundbreaking approach leveraging deep learning has significantly advanced the segmentation and analysis of point clouds, offering substantial promise for ScanToHBIM methodologies. This study demonstrates that artificial intelligence can effectively process large datasets from the Ducal Palace at Urbino, enhancing cognitive and management processes in cultural heritage (CH) research. By integrating semantic intelligence derived from HBIM models—an existing digital heritage—the research establishes a cyclical pathway where AI transforms specialized knowledge into collective intelligence, serving as a powerful tool for heritage preservation. The findings underscore the potential of this methodology to accelerate and improve the accuracy of architectural analysis and modeling in UNESCO World Heritage Sites.

Paolo Clini, Roberto Pierdicca, Ramona Quattrini, Emanuele Frontoni, & Romina Nespeca. (2021). Deep Learning for Point Clouds Classification in the Ducal Palace at Urbino. Representation Challenges. Augmented Reality and Artificial Intelligence in Cultural Heritage and Innovative Design Domain. https://doi.org/10.3280/oa-686.28