Motion prediction of wrist posture by neural networks using dynamic arm state
Taichi Watanabe, Takumi Aotani, Ryuta Ozawa · The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec) · 2023
Myoelectric prostheses are useful for improving the quality of life of amputees. Although current myoelectric analysis research has focused on increasing the number of identifiable grasp shapes, few studies have focused on estimating wrist posture. Therefore, in this study, we focus on the relationship between arm and wrist posture, and estimate the wrist posture in grasping behavior from upper arm and forearm posture continuously. Specifically, LSTM, a neural network capable of dealing with time series information, was used to estimate the wrist posture 0.01s ahead, using the arm posture up to 1s ago as input. As a result, we successfully surprized estimate the average angular error to less than 3.3 degrees.