FPGA-based Implementation of Random Forest Classifier for sEMG Signal Classification.

Area of Science:

• Biomedical Engineering
• Computer Science
• Machine Learning

Background:

• Developing accurate and efficient embedded systems for motion recognition is crucial for applications in prosthetics and human-computer interaction.
• Existing methods often face challenges in real-time processing and accuracy for complex movements.

Purpose of the Study:

• To propose and validate a novel embedded system for motion recognition.
• To combine hardware description language with Random Forests for enhanced performance.

Main Methods:

• A hardware description language was utilized to design the embedded system architecture.
• Random Forests theory was applied for the motion recognition algorithm.
• The Ninapro database DB2 was employed for training and testing the system.

Main Results:

• The developed embedded system demonstrated a high accuracy of 83.1% in movement prediction.
• The system's performance was validated on the Ninapro DB2 test dataset.

Conclusions:

• The proposed method effectively integrates hardware description language and Random Forests for motion recognition.
• The system shows significant potential for real-world applications requiring accurate movement prediction.