Advanced Hand Gesture Prediction Robust to Electrode Shift with an Arbitrary Angle

Area of Science:

• Biomedical Engineering
• Human-Computer Interaction
• Signal Processing

Background:

• Surface electromyogram (sEMG)-based sensing armbands offer convenient bioelectric signal acquisition.
• Electrode shift is a common issue causing classifier inaccuracies and requiring frequent recalibrations.

Purpose of the Study:

• To develop a hand gesture prediction system robust to arbitrary electrode shifts.
• To enable gesture recognition before the completion of sEMG signal collection.
• To implement rapid, simplified detection and correction of electrode shift at random angles.

Main Methods:

• Proposed a novel hand gesture prediction algorithm.
• Combined interpolated peak location and preset synchronous gesture for shift detection and correction.
• Developed a system for rapid electrode shift detection and correction at random angles.

Main Results:

• Achieved high precision in electrode shift detection.
• Demonstrated high accuracy in hand gesture prediction, even with electrode shift.
• Showcased the ability to predict gestures before signal collection completion.

Conclusions:

• The developed system offers robustness against electrode shift for sEMG-based gesture prediction.
• The approach brings advanced gesture prediction closer to practical applications by mitigating calibration issues.
• This research provides new insights into achieving electrode shift robustness in bioelectric signal processing.