Two-Stage Enhancement of Noisy and Reverberant Microphone Array Speech for Automatic Speech Recognition Systems Trained with Only Clean Speech.
Quandong WangSicheng WangFengpei GeChang Woo HanJaewon LeeLianghao GuoChin-Hui LeePublished in: ISCSLP (2018)
Keyphrases
- speech signal
- automatic speech recognition systems
- noisy speech
- speech recognition
- noisy environments
- automatic speech recognition
- vocal tract
- hidden markov models
- speaker identification
- linear prediction
- speech enhancement
- background noise
- additive noise
- spectral analysis
- adaptive filtering
- non stationary
- image processing
- signal processing
- computer vision
- spectral subtraction
- sound source
- neural network
- signal to noise ratio
- acoustic features
- feature extraction
- information retrieval