Electromyographic (EMG) signals provide crucial information about muscle activity and have applications in various medical fields. In recent years, artificial intelligence has played an increasingly important role in EMG signal analysis, driving the development of more efficient models for muscle movement classification. This study proposes a deep learning (DL)-based approach for classifying ten different right-hand movements using convolutional neural networks (CNNs) and transformers. The experimental results show an accuracy of 91.06% for a CNN model, 97.04% for a hybrid CNN-LSTM model, and 96.06% for a transformer-based model called PatchTST. These findings demonstrate that DL models can effectively classify muscle movements with high accuracy, particularly when combined with event-based windowing strategies for EMG signals.