Executive Summary : | The project aims to enable natural communication among participants in an authentic virtualized space using AI-supported navigation of autonomous drones, computer vision, and extended reality mechanisms. It comprises four interoperating modules: Acquisition, Autonomous navigation of drones, Virtualization, and Scene analysis. The project partners will select and test various acquisition devices for indoor and outdoor/mobile scenarios, including drones equipped with cameras for photogrammetric 3D reconstruction. Other devices will be tested for multi-sensor setups for optimal reconstruction.
FlyCamXR proposes an efficient approach to coordinate drones to ensure multiple-view geometry coverage. A swarm of drones will act as integrated acquisition devices, with the optimal flight pass pre-computed before flight. The drones will be capable of operating autonomously while following commands from a supervisor and in GPS-denied areas through sensor-based localization.
Virtualization, the main component of the proposed XR telecommunication pipeline, will increase the naturalness and authenticity of shared collaboration and replace virtual artificial environments. The project combines offline 3D reconstruction and real-time update and refinement using simultaneous localization and mapping SLAM approaches to improve the accuracy and quality of the reconstruction results.
FlyCamXR represents the first step towards a new approach to future remote interaction among people over the internet. The project aims to promote innovation potential at the product and business model level, generating future uses and applications for technologies developed that cannot be foreseen today. |