An EEG-Based BCI System to Facial Action Recognition

作者：Lin Jzau Sheng^*; Jiang Zih Yang

来源：Wireless Personal Communications, 2017, 94(3): 1579-1593.

DOI：10.1007/s11277-016-3700-3

• EEG
• BCI
• Wavelet
• SVM

Brain-computer interface (BCI) has created a new era in neuroscience. It has improved the life quality of severely disabled patients. It allows them to regain the power of executing will by their cognitive, expressive and affective brain activities. An electroencephalogram (EEG)-based BCI system with wireless manner was developed to extract EEG signals with Emotiv EPOC head set for recognizing the facial actions in this paper. The extracted feature vectors of EEG can be reduced by the Wavelet transform. Then the reduced EEG signals can then be clearly classified into six clusters by means of support vector machine algorithm with Gaussian kernel function. The better correct rates can be obtained by one-order wavelet transform than those got by three-order wavelet transform. In order to get real-time manner to control an electronic system smoothly, the sampling data have to be reduced. If time consumption is considered, we can choice the one-order wavelet transform with 32 samples. The experimental results showed a promising correct rate for the facial-action recognition through the proposed BCI system with real-time manner.

• 出版日期2017-6