Although it has been thought so far that neural activities in vivo appears signal changes outside the body, it would be difficult to obtain the characteristics of the activities from the signal changes.
Especially in EMG (ElectroMyoGram), pattern-recognized characteristic classification and extraction methods might be necessary to obtain the characteristics because it deals with many kinds of fuzzy signals. In this study, we propose a neural network system appropriate for bio-information processing.
The aim of the system is aid of the development of an EMG driven artificial hand and the interface of the VR system.
To Research Topics List of Chihara Lab.