Visualization in Motion: A Research Agenda and Two Evaluations

Abstract

Moving visualization & stationary viewer. Stationary visualization & moving viewer. Moving visualization & moving viewer.Fig. 1: Visualization scenarios that involve different types of relative movement between viewers and visualization: (a): 0 A.D. game characters with attached health meters, (b): an augmented basketball match from the tool Clipper CourtVision. (c): a walkable visualization of the general organization of scholars at ENAC in France [75], [76]. (d): an on-street bar chart that can be driven or walked by created by the Respect New Haven activist group. (e): a runner looking at her fitness data. (f): a person checking financial charts on her phone while walking to a meeting. Image permissions are listed in the acknowledgments.

A meta-framework for multi-label active learning based on deep reinforcement learning

“…We aim to apply RL-based methods to manage label layouts in dynamic 3D environments, an underexplored area with unique challenges. This work, similar to Yao et al [55], focuses on scenarios where the viewer is stationary but can rotate their viewpoint horizontally or vertically to observe multiple moving objects, such as watching sports in a stadium or monitoring environments using surveillance cameras. We constrain that each object is annotated by a single label and leave multiple-label or moving viewpoint scenarios for future research.…”

Evaluating View Management for Situated Visualization in Web‐based Handheld AR

“…These properties can act as a classifying framework for existing techniques, but can also be used as a generative lens for designing future situated views. Here we enumerate the basic properties of a visualization inhabiting a situated analytics environment [BFH01]: Position: The visualization's location in the environment; Size: Its geometric size in relation to the rest of the world; Transparency: The opacity of the visualization, which also incorporates its general geometry (i.e., some visualizations such as a 3D scatterplot are more sparse than a volume rendering); Priority: A relative priority for each individual visualization (potentially whether a visualization is selected or not); Orientation: The visualization's 3D rotation; Distance: Its distance from the viewer and other visualizations; Area of interest: The area (often a 3D volume) from which to optimally view the visualization (i.e., the location of the user within which a visualization becomes relevant); Spatial relation to the surrounding world: The visualization's relation to real objects in the physical world (e.g., the proximity of a visualization of accident data to dangerous stairs); and Relative motion: Dynamic motion of the visualization with respect to the viewer, e.g., when one or both are moving [YBVI22]. …”

“…• Area of interest: The area (often a 3D volume) from which to optimally view the visualization (i.e., the location of the user within which a visualization becomes relevant); • Spatial relation to the surrounding world: The visualization's relation to real objects in the physical world (e.g., the proximity of a visualization of accident data to dangerous stairs); and • Relative motion: Dynamic motion of the visualization with respect to the viewer, e.g., when one or both are moving [YBVI22].…”

Designing for Visualization in Motion: Embedding Visualizations in Swimming Videos

“…Visual representations often take the form of text, inserted temporarily at fixed locations on the display. With the availability of computer vision methods to track athletes, however, it becomes possible also to create situated visualizations in motion [1] that move entirely with the athletes [2] or equipment (e.g., balls, cars, ...). In swimming races, for example, record lines move along the pool according to an invisible athlete swimming at average record speed [3]- [5]; or in basketball [6], [7] visualizations on player shot-probabilities are shown with bar or donut charts above players' heads.…”

“…This indicates that the general public may be open to the addition of sophisticated visual representations. (b) There is the first evidence that people can very accurately read quantities from simple moving visualizations [1]. (c) Working with sports videos is always difficult due to copyright constraints that may make sharing research results difficult.…”