Two objective measures for speech distortion and noise reduction evaluation of enhanced speech signals.
Huijun DingTan LeeIng Yann SoonPublished in: ISCSLP (2012)
Keyphrases
- speech enhancement
- noise reduction
- speech signal
- noisy environments
- objective measures
- signal to noise ratio
- speech recognition
- edge preserving
- automatic speech recognition
- vocal tract
- noise level
- edge detection
- hearing aids
- noise removal
- linear prediction
- wiener filter
- automatic speech recognition systems
- evaluation criteria
- non stationary
- single channel
- spectral subtraction
- additive noise
- speech quality
- hidden markov models
- sound source
- quality assessment
- smoothing algorithm
- prediction error
- user satisfaction
- speech sounds
- image segmentation