Videos

This video presents an overview of the EVOCATION project.

This work demonstatrates a novel approach for assisting users in 2D data exploration with an interactive lens. Our first key-contribution is a novel user-interface to jointly control lens and camera to support effective focus-and-context exploration.

In 2D Data exploration, this work exploits and extends the conecpt of data annotations to provide guidance in discovering interesting areas while navigating with the lens suggesting aanotations and displaying them without introducing clutter.

This work demonstrates the standardization of optical material behavior, particularly in case of standardizing 3D color managemeent.

This work demonstrates the interaction methods for light field displays by means of a theater model environment.

This work demonstrates the realistic physical camera motion for light field visualization.

This work demonstrates a robust feature-aware sparse mesh representation.

This work demonstrates the acquisition, reconstruction, and scalable exploration of paintings from retable of San Bernardino.

This work demonstrates the acquisition, reconstruction, and scalable exploration of Stele Di Nora.

This work demonstrates a real-time light-field 3d telepresence system. The system is comprised of the capture and display subsystem with a general purpose network in between them.

This video show the tele-collaboration techniques and methods that were developed in EVOCATION. It includes a volumetric capture system, a real-time light-field 3d telepresence system, and a system for scalable foveated rendering.

This pilot project showcases the efforts focused on utilizing a prototype mobile mapping system to capture indoor 3D environments and demonstrates its applicability in relevant use cases within the AEC domain. All collected data is prepared for further analysis and processed, aiming to enhance and extract valuable information with techniques such as depth prediction and completion, clutter removal, semantic segmentation for CAD models generation and point cloud visualization.

A comprehensive overview of the data acquisition phase, which includes data capture using a prototype mobile mapping system in indoor environments, data preparation and its delivery in compatible formats suitable for the requirements of subsequent experimental and innovative data processing pipelines and visualization techniques.

This video presents the results from processing the captured data, focusing on improving and extracting information from panoramic input. We report on a method for depth prediction and completion capable of efficient data fusion for generating representations matching the visual input size. Additionally, we illustrate a method for extracting room layouts from panoramic input, and techniques for clutter removal applied to room emptying.

In this video, we present methods to convert point cloud data into structured information. We focus on an automatic reconstruction process consisting of three main steps: point cloud cleaning (including downsampling and noise removal), extraction of key building structures and architectural elements (such as walls, doors, windows, floors, and ceilings), and export of an integrated CAD model.

Application of novel techniques for point cloud visualization, showing their applicability in the field of AEC. We focus, in particular, on scalable methods that enable interactive display of large point clouds using ray-tracing techniques.