Dimensionality reduction of sample covariance matrices by graph fourier transform for motor imagery brain-machine interface.
Toshihisa TanakaTakashi UeharaYuichi TanakaPublished in: SSP (2016)
Keyphrases
- fourier transform
- covariance matrices
- motor imagery
- dimensionality reduction
- brain computer interface
- covariance matrix
- signal processing
- eeg signals
- frequency domain
- eeg data
- maximum likelihood
- principal component analysis
- brain activity
- distance measure
- feature extraction
- gaussian mixture model
- pattern recognition
- vector space
- gaussian mixture
- sample size
- high dimensional data
- gaussian distribution
- low dimensional
- independent component analysis
- principal components
- feature vectors
- linear classifiers
- image processing
- computer vision
- multiresolution
- high dimensional
- feature set
- image data
- feature space
- multiscale
- face recognition