Vector Autoregressive Hierarchical Hidden Markov Models for Extracting Finger Movements Using Multichannel Surface EMG Signals
We present a novel computational technique intended for the robust and adaptable control of a multifunctional prosthetic hand using multichannel surface electromyography. The initial processing of the input data was oriented towards extracting relevant time domain features of the EMG signal. Following the feature calculation, a piecewise modeling of the multidimensional EMG feature dynamics using
