In a breakthrough that could revolutionize human-machine interaction, researchers have developed a novel neural network model that significantly improves the accuracy of continuous motion estimation using surface electromyographic (sEMG) signals. This advancement, led by Chuang Lin from the School of Information Science and Technology at Dalian Maritime University in China, opens up new avenues for more intuitive and natural human-computer interfaces, with promising implications for the maritime sector.
The study, published in the journal ‘Applied Sciences’ (translated from the original ‘Applied Sciences’), introduces a parallel network that combines a Convolutional Neural Network (CNN) with a multi-head attention mechanism and a Bidirectional Long Short-Term Memory (BiLSTM) network. This hybrid approach extracts and fuses multiple features from raw sEMG signals, leading to enhanced accuracy in estimating finger joint angles.
“Better continuous estimation accuracy can be achieved using multi-feature sEMG data,” Lin explained. The proposed model demonstrated an average accuracy of 0.87 ± 0.02, significantly outperforming other models like BiLSTM, CNN-Attention, and CNN-BiLSTM. This improvement in accuracy is crucial for applications requiring precise and continuous motion control.
For the maritime industry, the implications are substantial. Imagine sailors and engineers interacting with complex machinery using intuitive gestures, reducing the need for cumbersome controls and interfaces. This technology could enhance the operation of robotic arms on ships, enabling more precise and efficient tasks such as maintenance and repairs. Additionally, it could improve the usability of exoskeletons and prosthetics for maritime workers, providing them with more natural and responsive control over their assistive devices.
The model’s stability and accuracy also make it suitable for training simulations and virtual reality applications in the maritime sector. Crew members could benefit from more immersive and realistic training environments, improving their skills and preparedness for real-world scenarios.
“This model realizes more natural and accurate human-computer interaction,” Lin noted. The integration of CNN, multi-head attention mechanism, and BiLSTM networks offers a robust solution for continuous motion estimation, paving the way for more intuitive and efficient human-machine interfaces in various maritime applications.
As the technology continues to evolve, the maritime industry stands to gain significantly from these advancements, enhancing both operational efficiency and worker safety. The research by Lin and his team represents a significant step forward in this exciting field, with the potential to transform how we interact with machines in the maritime environment.