Chenkai Zhang, Ruochen Cao, Andrew Cunningham, James A. Walsh
IEEE Conference on Virtual Reality and 3D User Interfaces (IEEE VR), 2024
[ Paper | Video ]
The Unity project for the paper 'Whatever could be, could be: Visualizing Future Movement Predictions'.
- Unity 2021.3.24f1 or later
- Google Research Football (using kaggle-environments to generate data)
- Soccer Players & Stadiums Pack (3D football environment)
- Oculus XR Plugin (Unity Package Manager)
- Newtonsoft Json (Unity Package Manager)
- Oculus Quest Pro (compatible eye-tracking)
As technology grants us superhuman powers, looking into what the future may hold is no longer science fiction. Artificial Intelligence and Mixed Reality technologies can allow users to see what the future may hold. In this paper, we present our work evaluating visualizations of future predictions in the Football domain. We explore the problem space, examining what a future may be. Three visualizations—2 Arrow Lines, 5 Arrow Lines, and Heatmap—are introduced as representations that show both individual predictions of movement (2 Arrow Lines and 5 Arrow Lines) and more generalized predictions (Heatmap). Whilst football is used as an example domain in this work, the visualizations and findings aim to generalize to other scenarios that contain trajectory information. Two VR studies (2 × n = 24) examined the visualizations in both simple/complex, timed/non-timed, and short/long-range viewing situations. Results show Heatmap as the most effective and preferred by the vast majority of participants. Findings offer insights into future visualization, serving as visual heuristics beyond the realm of sports and into the real world.
