Estimation of the excitation variances of speech and noise AR-models for enhanced speech coding.
Marcin KuropatwinskiW. Bastiaan KleijnPublished in: ICASSP (2001)
Keyphrases
- speech recognition
- noisy environments
- automatic speech recognition
- linear prediction
- augmented reality
- probabilistic model
- recognition engine
- speech enhancement
- endpoint detection
- noise reduction
- parameter estimation
- coding scheme
- speech synthesis
- statistical models
- noise level
- missing data
- image processing
- audio visual
- acoustic models
- autoregressive
- text to speech
- finite state transducers
- signal subspace
- broadcast news
- spoken language
- speech signal
- monte carlo simulation
- signal to noise ratio
- language model
- pattern recognition
- multiscale