Delayless speech enhancement with a virtual zero-phase response using a prediction of periodic signal components.
Kristian Timm AndersenThomas Bo ElmedybMarc MoonenPublished in: ICASSP (2015)
Keyphrases
- speech enhancement
- single channel
- noisy environments
- linear prediction
- wiener filter
- multi channel
- noise reduction
- frequency domain
- signal processing
- noisy speech
- speech signal
- non stationary
- signal to noise ratio
- prediction error
- background noise
- independent component analysis
- augmented reality
- speech recognition
- additive noise
- lossless compression
- image coding
- high frequency
- vocal tract
- feature vectors