Estimation of speech recognition performance in noisy and reverberant environments using PESQ score and acoustic parameters.
Takahiro FukumoriMasato NakayamaTakanobu NishiuraYoichi YamashitaPublished in: APSIPA (2013)
Keyphrases
- speech signal
- speech recognition
- noisy environments
- acoustic features
- speech recognition systems
- speaker independent
- noisy speech
- speech quality
- speech recognizers
- sound source
- automatic speech recognition
- hidden markov models
- speaker identification
- background noise
- speech recognizer
- vocal tract
- acoustic models
- language model
- adaptive filtering
- mel frequency cepstral coefficients
- speech enhancement
- spectral analysis
- pattern recognition
- speech processing
- speech synthesis
- speaker dependent
- non stationary
- speech retrieval
- maximum likelihood
- speaker verification
- speaker recognition
- information retrieval
- denoising
- multiscale
- multimedia