HandMatic: Wireless Robotic Arm Control Using Real-Time Gesture Recognition for Seamless Human-Machine Interaction

Abstract

The HandMatic system presents a groundbreaking approach to wireless robotic arm control through real-time gesture recognition. This innovative system combines a Hybrid CNN-BiLSTM architecture with a Transformer-based attention mechanism to overcome limitations in accuracy, latency, and robustness in human-machine interaction. Gesture data acquired using multimodal sensors, including IMUs, EMG, and vision-based systems, providing adaptability for various applications. Preprocessing techniques such as noise filtering and spatiotemporal feature extraction improve data quality, while inverse kinematics ensures precise robotic arm control. The proposed model achieves an impressive average accuracy of 94.7%, outperforming nine existing models. It also demonstrates resilience under noisy conditions, achieving 93.8% accuracy with 10% noise, alongside a low latency of 15ms and energy efficiency of 3.5 W. These attributes make the system ideal for real-time applications in healthcare, manufacturing, and assistive robotics. By seamlessly integrating advanced gesture recognition with robotic arm control, HandMatic provides a reliable and adaptive solution for enhanced human-machine collaboration. Its hybrid deep learning architecture establishes a new standard for real-time systems, demonstrating the potential for further developments in intuitive and interactive technology.