Program

Our approach utilizes a learning framework combined with NeRF representation to render novel views from limited source images in medical environments, specifically open surgical procedures. By incorporating automatic occlusion mask estimation to filter obscured pixels during training, we overcome challenges posed by occlusions and camera constraints in operating rooms. We validate our approach through comparisons using real-world surgical videos and demonstrate improved rendering quality and practical effectiveness in handling occlusions. 

Interaction in immersive VR can be enhanced by merging physical objects with virtual stimuli, thus creating extended reality scenarios. Effective solutions rely on the correct detection and tracking of the 6DOF pose of the objects and the accurate 3D reconstruction of the physical objects. Here, we focus on the latter aspect, analyzing different 3D reconstruction techniques to create and segment the meshes necessary to merge the physical objects with virtual counterparts. We consider six 3D reconstruction software and six different furnished rooms, analyzing the time necessary to create the 3D mesh from scratch, the reconstruction success rate, and the percentage of volume shift between the real objects and the reconstructed ones. 

In large-scale retail environments, the use of a digital twin - a virtual model of a physical store - can enhance efficiency and decision-making. However, creating and maintaining these digital twins is labor-intensive due to manual data entry, product tracking, and discrepancies between blueprints and actual store layouts. This paper proposes an AI-driven solution for automated data collection and updating for digital twinning. The framework enables real-time product recognition and location detection using head-mounted displays and smartphones. An AR feature allows for immediate data verification. The system's effectiveness was confirmed with an iOS wayfinding app, achieving high user satisfaction and accuracy. 

This work compares four possible designs of Augmented Reality (AR) interfaces for passengers of an Autonomous Aerial Vehicle (AAV) envisioned as an air taxi in the Urban Air Mobility (UAM) context. The four designs were evaluated and compared through a video-based study considering two potentially influential factors: flight phases (namely takeoff, cruise, and landing) and visibility conditions (i.e. clear daylight, night, and foggy). Dimensions included in the analysis were perceived safety, anxiety, situational awareness, cognitive workload, trust, predictability, and preference. The results showed that preferred interface by the passengers may vary depending on the considered combination of the two factors.