A new cascaded spectral subtraction approach for binaural speech dereverberation and its application in source separation.
Muhammad Salman KhanSyed Mohsen NaqviJonathon A. ChambersPublished in: ICASSP (2013)
Keyphrases
- source separation
- spectral subtraction
- speech signal
- blind source separation
- sound source
- single channel
- speech enhancement
- speech recognition
- audio features
- automatic speech recognition
- speaker identification
- vocal tract
- multi channel
- hybrid method
- non stationary
- noisy environments
- independent component analysis
- linear prediction
- hidden markov models
- additive noise
- pattern recognition
- frequency domain
- language model
- denoising
- transfer function
- noisy images
- multiscale