Expanding Touch Interaction Capabilities for Smart-rings: An Exploration of Continual Slide and Microroll Gestures

Abstract

As smart-rings emerge in both research and commercial markets, their limited physical size remains to restrict the interaction potential and input vocabulary possible. Thus, focusing on touch interaction for its natural and preferred input potential, this early work explores the combination of slide and microroll gestures performed by the thumb in continual motion on a smart-ring’s touch capacitive surface. We first capture over 3000 slide and microroll gesture instances, extract features, and generate and test machine learning models that are able to discern the slide and microroll gestures within the same touch instance. Through the use of 18 features, our Random Forest model provides a 92.4% accuracy. We conclude with demonstrations of potential applications utilizing continual slide and microroll gestures, and a short discussion which provides future research directions stemming from the positive results obtained from this preliminary work.

In Summary

In this work, we enable the ability for continual slide and roll gesturing on a smartring. Touch interaction on a smartring is greatly hindered by the small interaction space available. Yet importantly, touch interaction is a common and natural interaction modality that we should both utilize and take advantage of.

Key Findings

• Utilization of unique data capturing and processing steps to aid in simple data capture and model generation
• Real-time classification of location independent slides and microrolls within the same touch instance
  • A random forest model with 18 features (highest ranked feature was pressure) achieved a 92.4% recognition rate
• Demonstration of two potential useage scenarios of continual slide and micoroll gestures

In More Detail

Please review our late breaking work (linked above) for study details, methodologies, and complete results.