SVD-ICA: A new tool to enhance the separation between signal and noise subspaces.
Valeriu D. VrabieJérôme I. MarsPublished in: EUSIPCO (2002)
Keyphrases
- source signals
- independent component analysis
- principal component analysis
- blind source separation
- signal processing
- singular value decomposition
- random noise
- independent components
- blind separation
- additive noise
- blind signal separation
- white noise
- source separation
- received signal
- low signal to noise ratio
- signal detection
- dimensionality reduction
- single channel
- low snr
- face recognition
- wide band
- high frequency
- power spectra
- non stationary
- additive gaussian noise
- denoising methods
- denoising
- lower dimensional
- signal to noise ratio
- low frequency
- data points
- stochastic resonance
- original signal
- low dimensional
- response function
- frequency domain
- high dimensional
- noise variance
- noisy data
- statistically independent
- biomedical signals
- high dimensional data
- input data
- noise model
- power spectrum
- image processing
- short time fourier transform
- direct sequence spread spectrum
- feature extraction
- nonlinear functions
- multiplicative noise
- impulse response
- negative matrix factorization
- noise removal
- nonlinear filters
- pixel intensities