Designing a 3D Gestural Interface to Support User Interaction with Time-Oriented Data as Immersive 3D Radar Chart

TL;DR

This paper addresses the challenge of designing intuitive 3D gestural interfaces for immersive analytics to explore time-oriented data. It presents a 3D gestural UI based on hand-based grasping and gestural commands applied to 3D Radar Charts, implemented in VR with Leap Motion and HTC Vive. The authors detail a task-oriented design, implement features (travel, time-event selection, zoom, sort, filter, pause, reset), and provide an empirical evaluation with 12 participants showing good usability and engagement. They discuss lessons on hand detection, precision, and intent inference, and outline future work including robustness improvements and longitudinal studies.

Abstract

The design of intuitive three-dimensional user interfaces is vital for interaction in virtual reality, allowing to effectively close the loop between a human user and the virtual environment. The utilization of 3D gestural input allows for useful hand interaction with virtual content by directly grasping visible objects, or through invisible gestural commands that are associated with corresponding features in the immersive 3D space. The design of such interfaces remains complex and challenging. In this article, we present a design approach for a three-dimensional user interface using 3D gestural input with the aim to facilitate user interaction within the context of Immersive Analytics. Based on a scenario of exploring time-oriented data in immersive virtual reality using 3D Radar Charts, we implemented a rich set of features that is closely aligned with relevant 3D interaction techniques, data analysis tasks, and aspects of hand posture comfort. We conducted an empirical evaluation (n=12), featuring a series of representative tasks to evaluate the developed user interface design prototype. The results, based on questionnaires, observations, and interviews, indicate good usability and an engaging user experience. We are able to reflect on the implemented hand-based grasping and gestural command techniques, identifying aspects for improvement in regard to hand detection and precision as well as emphasizing a prototype's ability to infer user intent for better prevention of unintentional gestures.

Figures (6)

• Figure 1: Excerpt of a VE from the VR user's field of view, and interacting through 3D gestural input with a 3D Radar Chart (incl. Time Slice and juxtaposed Information Window).
• Figure 2: Illustration of the 3D UI design approach.
• Figure 3: Feature overview of the implemented 3D gestural interface (see \ref{['tab:3duioverview']}): (a) Travel; (b) Mode Toggle; (c) Rotation; (d) Time Event Selection; (e) Data Variable Filter (top) and Sort (bottom); (f) Time Range Selection; (g) Zoom (in); (h) Zoom (out); (i) Reset; (j) Pause (top) / Resume (bottom). See also video demonstration (URL provided in Footnote \ref{['footnote:demo']}).
• Figure 4: Applied hand posture comfort configurations, adapted from the recommendations by rempel2014tdo. To facilitate cross-referencing to their work rempel2014tdo, we apply here the same label coding for convenience: 1--12 = hand posture example; c = comfortable; u = uncomfortable.
• Figure 5: Immersed participants during their task completion, wearing an HMD and interacting in the VE with the designed VR prototype as described throughout Section \ref{['sec:vrprototypeand3duidesign']}.