TL;DR
This paper argues that 3D user interface (UI) research has hit a plateau in discovering novel interaction techniques. To advance the field and improve real-world usability, the authors propose shifting focus from general techniques to more specific, "flavored," implementable, and context-aware approaches, particularly for emerging display technologies.
What Problem Does This Paper Address?
The paper addresses a perceived stagnation in 3D user interface (UI) research. While significant progress has been made in identifying fundamental 3D interaction tasks (like travel, selection, manipulation) and developing basic techniques, the authors observe a slowdown in the discovery of truly novel metaphors and techniques in recent years. Crucially, despite this foundational work, the usability of 3D UIs in many real-world applications remains disappointingly low. This suggests that simply creating more general techniques isn't the path forward. The paper seeks to identify new research directions that can bridge the gap between theoretical advancements and practical, usable 3D UIs.
Methods
The authors propose a research agenda centered around four key concepts:
- Specificity: Moving away from "one-size-fits-all" general techniques towards designs tailored for specific applications, domains, tasks, devices, or user groups. This acknowledges that a technique optimized for one context might be suboptimal in another.
- Flavors: Enhancing existing fundamental techniques by adding features, complexity, or subtle modifications to address specific usability issues or improve performance. This is about refining and adapting, rather than reinventing.
- Implementation: Recognizing and addressing the significant challenges in developing robust and efficient 3D UIs. This includes the need for better tools, libraries, and frameworks that support rapid prototyping and complex interaction design.
- Emerging Technologies: Investigating how 3D interaction principles and techniques can be applied to or adapted for new display and computing paradigms, such as large-screen displays and ubiquitous computing, even when the displayed content isn't inherently 3D.
The paper illustrates these concepts through case studies from the authors' own research group, showcasing projects that embody these new directions.
Datasets
Not discussed in the paper. The paper focuses on conceptual arguments and case studies of research projects, rather than empirical evaluations on specific datasets.
Key Insights and Ideas
- The "3D Interaction Boom" is Over: The paper presents evidence (publication dates of techniques) suggesting that the period of rapid discovery of fundamental 3D interaction techniques has largely concluded, and the focus needs to shift.
- Generality is a Double-Edged Sword: While general techniques are broadly applicable, their lack of specificity can lead to usability problems in real-world applications. Tailoring techniques to specific contexts is crucial for improving user experience.
- Implementation is a Bottleneck: The difficulty in building and prototyping complex 3D UIs with current tools is a significant barrier to research and adoption. Developing better implementation frameworks is essential.
- Beyond VR/AR: The principles of 3D interaction are relevant and potentially transformative for a wider range of technologies, including large displays, even for 2D content.
Key Results
The paper doesn't present a single set of quantitative results but rather uses case studies to demonstrate the potential of its proposed directions:
- Domain-Specificity (AEC Domain): Developing "cloning" techniques for the Architecture, Engineering, and Construction (AEC) domain led to significant usability gains by addressing a previously unconsidered task specific to that domain. This approach proved more effective than trying to adapt generic techniques.
- Task-Specificity (Resizing): Novel techniques for resizing virtual objects (PORT and Gaze-Hand) were found to be significantly faster and preferred over traditional widget-based methods, demonstrating the benefit of task-specific design.
- Display-Specificity (IRVEs): Research on information layout for Information-Rich Virtual Environments (IRVEs) showed that techniques effective on desktop displays did not necessarily translate well to large displays, highlighting the need for display-specific design considerations.
- Flavors (SSWIM): Enhancing the World-in-Miniature (WIM) technique with "panning and zooming" (SSWIM) improved accuracy in travel tasks without a significant drop in efficiency or user preference, suggesting that adding complexity can be beneficial.
- Implementation (CHASM): The development of the CHASM framework, based on state machines, shows promise for creating more understandable, robust, and reusable 3D UI implementations, though it has a steep learning curve.
- Emerging Technologies (Large Displays): Exploration of 3D input for large displays indicated that techniques like using a 3D tracker for navigation can be effective, especially on larger screens, and are preferred over traditional mouse input.
Limitations and Future Work
The authors acknowledge that their proposed directions are not exhaustive and that other areas, such as embodiment in HCI, are also relevant. They emphasize that the benefits of specificity need to be weighed against the increased development cost. Future work should focus on:
- Understanding when the advantages of specific 3D UIs outweigh the costs.
- Developing better tools and frameworks for rapid prototyping and implementation of 3D UIs.
- Investigating how existing 3D interaction techniques can be adapted or redesigned for emerging technologies like large displays.
- Exploring the integration of disparate 3D interaction techniques into cohesive user interfaces.
- Further research into the application of 3D interaction to technologies beyond VR and AR.
Summary updated