Reverberant audio source separation using partially pre-trained nonnegative matrix factorization.
Mahmoud FakhryPiergiorgio SvaizerMaurizio OmologoPublished in: IWAENC (2014)
Keyphrases
- nonnegative matrix factorization
- audio features
- negative matrix factorization
- matrix factorization
- audio visual
- data representation
- least squares
- multimedia
- visual features
- signal processing
- spectral clustering
- feature set
- principal component analysis
- speech signal
- document clustering
- sparse representation
- music information retrieval
- original data
- visual data
- data analysis
- pattern recognition
- training data
- data mining