Konversatorium on Friday, 3. December 2021, 10:30

Seminar room FAV05 (Favoritenstraße 9, Stiege 1, 5th floor)
Slots for talks still available! Please contact the KV administration.
Elitza Vasileva (VRVis)
20 + 20
Harald Steinlechner

Immersive systems like Virtual Reality (VR) are used in different application domains and for multiple purposes, including visual analytics. VR systems are capable of rendering realistic images and 3D scenes and creating the feeling of full immersion and presence. In this work, we develop an interactive visualization tool in VR to support domain experts in studying the properties and features of asteroid impact events for defense purposes. We use the time-dependent multivariate impact simulation data. The implementation requirements are formulated together with domain experts in the form of tasks and represent the main features that the system should include. As a result, the system incorporates a 3D point cloud visualization to illustrate the impact and the data structure and various exploration tools to analyze and examine its properties. The central tool in the system is called a probe, allowing to measure the characteristics of different regions, compare them, and observe state changes during simulation time. While effective exploration is the primary goal of our system, interactivity is a significant factor contributing to achieving a smooth and natural experience. Therefore, we provide various grasping and navigation techniques to support an intuitive and effortless system interaction. As the selection of exploration tools is essential for the domain experts and solving their tasks, we first evaluate our system with them to answer whether the system is providing the necessary features and is fulfilling their requirements. Another important measurement is the interactivity and usability of our system, which we evaluate through a user study. As we show in our evaluation experiments, our VR system eases the exploration process for scientists. It supports them in finding out new and previously undiscovered properties, patterns, and trends of the data.