FingerText: Exploring and Optimizing Performance for Wearable, Mobile and One-Handed Typing

Title of the Paper

FingerText: Exploring and Optimizing Performance for Wearable, Mobile, and One-Handed Typing

Paper Information

• Domain: Human-Computer Interaction and Input Performance Optimization for Wearable Devices
• Keywords: Wearable input, nail touch sensor, one-handed input, mobile input, keyboard layout optimization, human-computer interaction, smart nail keyboard, input performance while walking

Research Background and Problem

• Problem or Challenge: Traditional input methods for wearable devices face challenges such as limited space, low precision, insufficient visual feedback, and poor input performance in mobile environments. Existing input technologies are mostly designed for stationary use and fail to adequately address the challenges of mobile conditions.
• Significance: With the widespread use of smart devices in mobile scenarios (e.g., walking or busy states), efficient one-handed input is crucial for improving user experience, reducing distractions, and mitigating potential safety risks.
• Research Motivation: The authors propose that a one-handed input method based on nail touch sensors is more resistant to interference in mobile environments. Additionally, optimizing the keyboard layout can further enhance speed, comfort, and accuracy.
• Related Work: Previous studies mainly focused on two-handed devices or fingertip input methods, but their speed and accuracy were limited or reliant on fixed environments. There has been no systematic exploration of one-handed input performance under real mobile conditions.

Solution

• Proposed Method: Using nail touch sensors for one-handed input and designing two keyboard layouts through multi-objective optimization: F5 (five-key layout) and F10 (ten-key layout). The optimization aims to balance speed, accuracy, comfort, and unambiguity in text input.
• Innovations: By combining input performance data under mobile conditions, assessments of touch comfort, and a bigram input model, the study conducts four-dimensional objective optimization to generate the optimal keyboard layouts.
• Key Technologies and Implementation Steps:
  1. Experimental Platform Setup: Nail-mounted capacitive touch sensors were used to collect user input data. The sensors can detect multiple touchpoints on each nail surface.
  2. Data Collection and Analysis: Input speed and accuracy were compared between sitting and walking conditions, revealing minimal impact of mobility on input performance.
  3. Keyboard Layout Design and Optimization: A genetic algorithm (NSGA-II) was used to optimize metrics such as speed, accuracy, comfort, and word-level input unambiguity, resulting in the selection of F5 and F10 layouts.
  4. Performance Evaluation of Optimized Keyboards: The optimized layouts were compared with the QWERTY baseline keyboard to validate performance improvements.

Research Outcomes

• Specific Results:
  • The F5 layout improved input speed by 9.47% per minute and reduced error rate by 23.68%.
  • The F10 layout improved input speed by 10.45% per minute and reduced error rate by 39.44%.
  • Compared to the QWERTY baseline keyboard, the optimized layouts significantly enhanced comfort and text input performance under mobile conditions.
• Advantages:
  • The design supports fast, accurate, comfortable, and unambiguous text input.
  • Performance remained largely unaffected while walking.
• Experimental and Evaluation Results: Input speed and accuracy remained stable during long-term testing, indicating that the optimized keyboard layouts are suitable for real-world mobile scenarios.
• Limitations and Future Directions:
  • Limitations:
    1. Other mobile input challenges (e.g., use in vehicles) were not considered.
    2. Multi-round learning curve tests were not conducted to verify performance under fully familiarized conditions.
    3. Comparisons were limited to the QWERTY layout, without testing other potential layouts such as alphabetical order.
  • Future Directions:
    1. Incorporate more finger regions or touchpoints to enrich input methods.
    2. Extend optimization algorithms to explore more multidimensional optimal solutions.
    3. Design and test more complex input tasks (e.g., integrating auto-completion and text suggestion features).

Conclusion

This study demonstrates the potential of nail touch sensors for one-handed input under mobile conditions. By optimizing layout design, it improves speed and accuracy, providing valuable references for future text input solutions on wearable devices. The research also introduces new directions for designing mobile-friendly input systems, particularly for enhancing user experience in real-world mobile scenarios.