Learning Rate Optimization for Enhanced Hand Gesture Recognition using Google Teachable Machine

Developing efficient sign language recognition systems using wearable devices is a major challenge in Machine Learning. One obstacle is effectively translating gestures based on sensor data. Traditional methods involve complex programming using data fusion and mapping techniques. To address this, we need emerging technologies that simplify gesture data processing while maintaining accuracy. This study explores an artificial intelligence approach for detecting Bahasa Melayu using a ready-to-use machine learning framework-Google Teachable Machine. By experimenting with these tools, the research aims to improve the simplicity and accuracy of hand gesture detection. The study also investigates the impact of the learning rate, an important parameter in machine learning algorithms, on system performance, providing insights for optimizing gesture detection. The results of our study emphasize the significance of thoughtfully choosing the learning rate for successful model training. This underscores the importance of finding the optimal learning rate to ensure effective training, regardless of the specific machine learning framework employed.