Towards a hardware realization of time-frequency source separation of speech.
Naomi HarteNiall P. HurleyConor FearonScott RickardPublished in: ECCTD (2005)
Keyphrases
- source separation
- audio features
- blind source separation
- independent component analysis
- speech signal
- speech recognition
- denoising
- single channel
- frequency domain
- audio visual
- signal processing
- low level
- temporal structure
- wavelet transform
- music information retrieval
- visual features
- audio signal
- non stationary
- feature set
- speaker identification
- image processing
- image restoration
- multiresolution