The implementation methods of Chinese phonetic brain-machine interface based on DIVA model

摘要

DIVA(Direction Into Velocities Articulators) model is a mathematical model which describes the effect of the brain involved in speech production and speech understanding area that plays a role. It can do the simulation of pronunciation process. It has the guiding significance of the design for phonetic brain-machine interface system. The user only need to think about what kind of language that he wants to express. The phonetic brain-machine interface system can transform it into voice directly. The article aims at the design of Chinese phonetic brain-machine interface. First of all, gather the electro encephalo gram of the Chinese enunciators when they are speaking the Chinese pinyin pronunciation. Then combine the FMRI(functional magnetic resonance imaging) test that is associated with Chinese phonetic pronunciation. Use the method of common spatial patterns (CSP) to extract the electro encephalo gram features. Then do the classification of the features by using the Support Vector Machine. The result proves that this way has the high accuracy of the brain electrical signal recognition of the Chinese people's pinyin pronunciation. It can apply to the exploitation and design of the Chinese people's brain-computer interface.