Spatial Separation of Speech Signals Using Continuously-Variable Masks Estimated From Comparisons of Zero Crossings.
Hyung-Min ParkRichard M. SternPublished in: ICASSP (4) (2006)
Keyphrases
- speech signal
- zero crossing
- edge detection
- sound source
- speech recognition
- automatic speech recognition
- automatic speech recognition systems
- noisy images
- scale space
- noisy environments
- spatial information
- edge detector
- vocal tract
- non stationary
- wavelet transform
- blind source separation
- sound signals
- speech enhancement
- image features
- pattern recognition
- multiscale
- neural network
- gray scale images