Harmonic beamformers for speech enhancement and dereverberation in the time domain.
Jesper Rindom JensenSam Karimian-AzariMads Græsbøll ChristensenJacob BenestyPublished in: Speech Commun. (2020)
Keyphrases
- speech enhancement
- single channel
- noisy environments
- noise reduction
- frequency domain
- signal to noise ratio
- speech signal
- linear prediction
- smoothing algorithm
- speech recognition
- multi channel
- vocal tract
- wiener filter
- background noise
- edge preserving
- additive noise
- blurred images
- feature extraction
- independent component analysis
- image coding
- edge detection
- probabilistic model