Search Results

The current paper identifies an effective deep learning Convolutional Neural Networks (CNN) model which is used in accurate recognition of hand gesture with signals through Electromyography (EMG). The suggested system utilises the capability of CNNs to automatically learn and extract discriminating features of raw EMG signals without manual feature engineering. Signals captured through EMG on forearm muscles are then processed and passed to a multi-layered CNN architecture that is aimed at learning spatial and temporal dynamics related to a number of hand movements. According to the experimental outcomes, the model exhibits strong classification robustness rate and accuracy levels on various subjects and gesture classes exemplifying its applicative capability in real-time, including prosthetic device control, human-computer interaction, and rehabilitation-related scenarios. The suggested CNN-based system will be reliable, scalable, and user-friendly to the gesture recognition problem, providing evidence of the high potential of deep learning in biomedical signal processing.