Underdetermined blind separation of delayed sound sources in the frequency domain.
Pau BofillPublished in: Neurocomputing (2003)
Keyphrases
- blind separation
- frequency domain
- sound source
- convolutive mixtures
- blind source separation
- speech signal
- source signals
- single channel
- spatial domain
- cross correlation
- filter bank
- speech recognition
- independent component analysis
- low pass
- denoising
- feature extraction
- focus of attention
- audio features
- single image
- audio visual
- subband
- machine learning
- real environment
- non stationary
- image registration
- source images
- motion estimation
- probabilistic model
- hidden markov models