SNOUT: One-Handed use of Capacitive Touch Devices

1. SNOUT: One-Handed use of Capacitive Touch Devices Adam Zarek, Daniel Wigdor, Karan Singh University of Toronto

2. Problem

3. Exploring the solution space • Survey: • We presented 6 scenarios where conventional finger input was not possible, seeking input alternatives. • …wearing gloves, dirty or preoccupied hands. • 15 participants (13 male, 12 owning a touch device). • Results: • 86% of participants suggested the non-conventional appendages: nose, toe, elbow, knuckle at least once. • the nose provided the broadest coverage across participants, 60% suggesting it in at least one scenario.

4. Touch-free Alternatives

5. Nose Pilot Study • How accurate is nose-based interaction? • 8 participants (7 male, all familiar with touch devices) • Results: Mean miss distance from the target was 0.43 times the target size Target size should be scaled by 1.5x the target width 4/5mm 8/11mm 12/15mm

6. Design Principles from pilot study • Avoid sliding since will dirty the screen • UI elements must be “large enough” • Focusing on what the nose is currently touching induces eye-fatigue • Avoid repeated up-down motions

7. Design Goals • Minimize nose taps • Minimize nose sliding • Avoid inducing eye-fatigue • Preserve existing UI layouts • Mitigate interaction errors

8. SNOUT Design • Selection • Text Entry • Continuous parameter specification

9. SNOUT selection • Increase target size by 1.5x (pilot study) • Reduce reliance on focused visual feedback • Solution: Two-stage color-based selection • Pre-processing: • split UI into Voronoi regions • apply cyclic color patterns • step 1: aim for the selection target flooding the periphery with the color of the currently selected region • step 2: slide nose on screen until the color of the desired selection is on the periphery, then disengage

10. SNOUT selection

11. SNOUT text entry • Small keyboard • Single appendage • No visual focus on selection • => repeated key selection cumbersome • Augment color-based selection • with speech recognition • Tapping a text box launches • speech recognition service • Corrections are made using • color-based selection

12. SNOUT text entry

13. SNOUT continuous parameters • No visual focus • Minimize sliding • => direct touch manipulation is difficult • Control the parameter by tilting the device • Enter tilt mode via touch-and-tap gesture • Exit tilt mode via hardware volume buttons • Touch-and-tap: hold the device to your nose and then tap on the back of the device

14. SNOUT continuous parameters

15. Usability Study • Method: • 3 custom applications that use our interaction techniques • Asked to accomplish a set of tasks within each application • Before/after survey about willingness to use different body parts as input methods • Participants: • 12 participants (9 male, 22-35, all with mobile experience)

16. Applications • Application Launcher: selection and scrolling • Launch 5 applications at different screen locations including scrolling • Notepad: speech to text and keypad selection • Text input of at least 10 characters and less than 3 characters. Text input of varying length using both text entry methods. • Map Browser: scrolling and zooming • Map navigation exercise, viewing a round trip of checkpoints in Cambridge MA, Mexico City, Tokyo.

17. Qualitative Results • Peripheral color based selection • Generally successful, works well in practice • Overhead of keeping target color in memory • Works better when focus is on color than tip of nose • Text entry • Generally preference for text to speech • Direct selection useable for short messages • Additional support for word-level selection desired • Continuous parameters • Touch+tap is an easily performed gesture • Works as well/poorly as tilt based parameter control • Some users attempted to use typical sliding motions

18. Quantitative Results

19. Message • Peripheral feedback has general UI potential • Potential as accessability interface.

20. Thank you!