Single Channel Speech Music Separation Using Nonnegative Matrix Factorization with Sliding Windows and Spectral Masks.
Emad M. GraisHakan ErdoganPublished in: INTERSPEECH (2011)
Keyphrases
- nonnegative matrix factorization
- sliding window
- single channel
- sound source
- speech signal
- multi channel
- data streams
- audio visual
- negative matrix factorization
- matrix factorization
- data representation
- speech recognition
- spectral clustering
- least squares
- principal component analysis
- independent component analysis
- edge detection
- frequency domain
- prior information
- wiener filter
- non stationary
- machine learning
- document clustering
- image segmentation
- linear combination
- dimensionality reduction
- image processing
- denoising
- objective function
- feature extraction