Observation-based design methods for gestural user interfaces

  • Authors:
  • David L. Akers

  • Affiliations:
  • Stanford University, Stanford, CA

  • Venue:
  • CHI '07 Extended Abstracts on Human Factors in Computing Systems
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

The design of gestural user interfaces is uniquely challenging because the input is freeform, personal, and often carries subconscious meanings that are domain-specific and difficult to articulate. These features suggest an approach of observation-based design: learning from what people do, rather than relying on what they say. To facilitate observation-based design, this dissertation is exploring two design methods: gesture brainstorming, a Wizard of Oz method for early prototyping of new interfaces, and gesture log analysis, a machine learning-based log analysis method for improving existing interfaces. These design methods will be tested by applying them to two gestural interfaces: a 3D pathway selection interface (CINCH, see Figure 1), and a 3D modeler (Google SketchUp). Experience with CINCH already suggests the utility of observation-based design, while work on Google SketchUp is anticipated to begin this summer. These test cases should inform observation-based design for gestural user interfaces in general.