Bevel or Not: Identifying the Potential of Bevels for Touch Input Accuracy on Ring Device

Paper Title

Bevel or Not: Identifying the Potential of Bevels for Touch Input Accuracy on Ring Device

Publication Info

• Topic area: Enhancing touch input accuracy and expressiveness on smart ring devices.
• Keywords: Smart rings, touch input, beveled surfaces, fat-finger problem, wearable devices, gesture classification, input accuracy, eyes-free interaction, capacitive sensing, ring design.

Background and Problem

• Problem / challenge: Smart rings have limited touch input accuracy and expressiveness due to their small surface area and susceptibility to fat-finger effects, especially on narrow bands.
• Significance: Improving touch input accuracy and expressiveness is crucial for enabling smart rings to support advanced interaction scenarios while maintaining wearability.
• Motivation and related work: Prior studies on smart rings and edge-based interactions have explored various touch input techniques but have not systematically addressed the fat-finger problem or evaluated the potential of beveled surfaces for touch input on rings. This paper addresses these gaps.

Solution

• Proposed approach: The use of beveled surfaces, in addition to the flat outer surface, to improve touch input accuracy and enable expressive gestures on ring devices.
• Novelty:
  1. Systematic investigation of the fat-finger problem in edge touch input across different ring shapes and band widths.
  2. Identification of beveled surfaces as a method to enhance input accuracy and expressiveness on narrow rings.
  3. Evaluation of nine touch gestures on beveled and flat rings under sighted and eyes-free conditions.
• Procedure and key techniques:
  • Fabrication of 3D-printed ring prototypes with flat, beveled, and rounded shapes.
  • Capacitive touch sensing system for detecting touch gestures.
  • Two user studies: (1) examining surface targeting accuracy across ring shapes and band widths, and (2) evaluating gesture classification accuracy and false positive rates for beveled and flat rings.

Results

• Concrete findings:
  • Beveled rings achieved 92.9% (sighted) and 91.8% (eyes-free) gesture classification accuracy with FPR ≤1% (0.88% sighted, 1.00% eyes-free).
  • Flat rings had significantly lower classification accuracy (63.8% sighted, 62.9% eyes-free) and higher FPRs (4.41% sighted, 4.55% eyes-free).
  • Beveled and rounded rings significantly improved surface targeting accuracy over flat rings at 6 mm and 4 mm band widths.
• Advantage over baselines: Beveled rings outperformed flat rings in both surface targeting and gesture classification accuracy, particularly on narrow bands prone to fat-finger effects.
• Experiments / evaluation:
  • Study 1: 15 participants performed swipe gestures on rings with three shapes (flat, beveled, rounded) and three band widths (4 mm, 6 mm, 8 mm).
  • Study 2: 12 participants performed nine touch gestures on flat and beveled rings with a 6 mm band width under sighted and eyes-free conditions.
  • Metrics: Surface targeting accuracy, gesture classification accuracy, and false positive rates.
• Limitations and future work:
  • Limited to sparse electrode layouts; effect on dense electrode arrays remains unclear.
  • Experiments conducted under controlled conditions; real-world scenarios (e.g., walking) need further exploration.
  • Future work includes integrating electronics into self-contained rings and exploring advanced classification methods.

Summary

This paper investigates the potential of beveled surfaces to improve touch input accuracy and expressiveness on smart rings. Two studies demonstrate that beveled rings outperform flat rings in surface targeting accuracy and gesture classification, particularly on narrow bands where fat-finger effects are prevalent. Beveled rings achieved over 91% classification accuracy with FPR ≤1% under both sighted and eyes-free conditions. These findings highlight the promise of beveled designs for enabling precise and expressive touch input while preserving the wearability of smart rings. Future work will focus on real-world deployment and advanced classification techniques.