EEG based direct speech BCI system using a fusion of SMRT and MFCC/LPCC features with ANN classifier.
P. P. MiniTessamma ThomasR. GopikakumariPublished in: Biomed. Signal Process. Control. (2021)
Keyphrases
- mel frequency cepstral coefficients
- feature set
- spectral features
- speaker identification
- feature extraction
- brain computer interface
- speech signal
- extracted features
- feature level fusion
- speaker recognition
- speech recognition
- artificial neural networks
- feature space
- probabilistic neural network
- lexical features
- svm classifier
- classification method
- eeg data
- healthy subjects
- gaussian mixture model
- multiple features
- data fusion
- motor imagery
- feature vectors
- complementary features
- cepstral coefficients
- audio features
- acoustic features
- neural network
- eeg signals
- signal processing
- training data
- feature selection
- support vector machine
- classification accuracy
- low level
- training set
- structural features
- non stationary
- speaker verification
- principal component analysis
- active learning
- support vector
- learning algorithm