PortInput: Enabling Always-Available Micro-Gesture Input With Pressure Array Sensor.

Area of Science:

• Human-Computer Interaction
• Wearable Technology
• Sensor Technology

Background:

• Micro-gestures offer natural, private input but are often unreliable in real-world conditions.
• Existing methods struggle with environmental dependencies, limiting their practical use.

Purpose of the Study:

• To develop an environment-independent micro-gesture interaction system using portable pressure sensors.
• To create a robust, real-time system for accurate micro-gesture recognition and tracking.

Main Methods:

• Developed three portable prototypes (finger-cot, finger-worn, surface-based) with compact pressure array sensors.
• Proposed a deep learning model to classify 14 micro-gestures from temporal pressure patterns.
• Integrated components into the PortInput real-time interactive system.

Main Results:

• PortInput demonstrated robust micro-gesture tracking and detection in user studies with augmented reality (AR) head-mounted displays (HMDs).
• Achieved comparable accuracy to commercial devices, with faster input and lower perceived workload.
• Showed improved usability and user experience in both sitting and walking conditions.

Conclusions:

• PortInput provides efficient, robust, and comfortable micro-gesture interaction across diverse scenarios, including AR/VR, smart homes, and vehicles.
• The system functions effectively even in noisy or cluttered environments.
• This work lays the groundwork for integrating pressure sensors into portable devices for ubiquitous computing.