Frequency-domain maximum likelihood estimation for automatic speech recognition in additive and convolutive noises.
Yunxin ZhaoPublished in: IEEE Trans. Speech Audio Process. (2000)
Keyphrases
- maximum likelihood estimation
- frequency domain
- automatic speech recognition
- speech recognition
- spatial domain
- fourier transform
- maximum likelihood
- em algorithm
- speech signal
- hidden markov models
- parameter estimation
- broadcast news
- power spectrum
- noisy environments
- denoising
- feature extraction
- expectation maximization
- conversational speech
- probability distribution
- density function
- cross correlation
- subband
- discriminative training
- speaker identification
- power spectra
- image denoising
- machine learning
- language model
- least squares
- computer vision