Robust underdetermined blind audio source separation of sparse signals in the time-frequency domain.
Si-Mohamed Aziz SbaiAbdeldjalil Aïssa-El-BeyDominique PastorPublished in: ICASSP (2011)
Keyphrases
- mixing matrix
- source separation
- blind source separation
- independent component analysis
- audio features
- signal processing
- source signals
- single channel
- blind separation
- denoising
- temporal structure
- non stationary
- audio visual
- independent components
- multiscale
- sparse representation
- image processing
- multi channel
- factor analysis
- frequency domain
- visual features
- feature set
- pattern recognition
- multimedia
- music information retrieval
- audio signal
- wavelet transform
- high dimensional