EMG-based Real-time Hand Gesture Prediction
Development of a data-driven model for real-time complex hand gesture recognition using EMG signals collected from subjects.
Period: 2018.03 – 2021.12
Affiliation: Incheon National University, Bioelectronics Lab.
Summary: Collected electromyogram (EMG) signals from subjects via sensors and developed a data-driven model capable of recognizing complex hand gestures in real time.
Key Responsibilities:
- Designed a prospective study and obtained IRB approval based on literature review
- Recruited subjects and conducted experiments following the study protocol
- Designed bandpass and band-stop filters using LabVIEW for real-time signal noise removal
- Extracted handcrafted features per window size using moving window technique
- Developed EMG handcrafted feature-based machine learning models
- Performed comparative study via statistical analysis and selected the optimal model (ANN)
Outcome: 1 SCI publication, Grand Prize at paper presentation