Speech intelligibility improvement using convolutive blind source separation assisted by denoising algorithms.
Jedrzej KocinskiPublished in: Speech Commun. (2008)
Keyphrases
- blind source separation
- speech signal
- speech recognition
- denoising algorithm
- denoising
- noisy images
- independent component analysis
- image denoising
- non stationary
- blind separation
- noise model
- noise reduction
- hidden markov models
- noise level
- wavelet coefficients
- transform domain
- negative matrix factorization
- edge detection
- language model
- signal to noise ratio
- visual quality
- sparse representation
- total variation
- wavelet domain
- factor analysis
- natural images
- probabilistic model
- feature extraction