Maximum-likelihood-based cepstral inverse filtering for blind speech dereverberation.
Kshitiz KumarRichard M. SternPublished in: ICASSP (2010)
Keyphrases
- maximum likelihood
- speech recognition
- linear prediction
- cepstral features
- speaker adaptation
- speech signal
- speech synthesis
- adaptive filtering
- expectation maximization
- audio features
- hidden markov models
- em algorithm
- information filtering
- automatic speech recognition
- speech recognizer
- speaker verification
- audio visual
- additive noise
- wiener filtering
- log likelihood function
- vocal tract
- maximum a posteriori
- filtering algorithm
- cross correlation
- recognition engine
- language acquisition
- gaussian distribution
- spoken language
- text to speech
- language model
- speaker independent
- dialogue system
- broadcast news
- image filtering
- likelihood function
- feature extraction
- synthetic aperture radar
- image coding
- pattern recognition