A frequency-weighted Itakura spectral distortion measure and its application to speech recognition in noise.
Frank K. SoongMan Mohan SondhiPublished in: IEEE Trans. Acoust. Speech Signal Process. (1988)
Keyphrases
- speech recognition
- spectral subtraction
- distortion measure
- cepstral coefficients
- noisy environments
- vector quantization
- speech signal
- human perception
- language model
- hidden markov models
- automatic speech recognition
- rate distortion
- structural similarity
- image coding
- speech recognizer
- speech synthesis
- spatial domain
- speech recognition systems
- background noise
- pattern recognition
- vector quantizer
- bit rate
- speaker identification
- linear prediction
- probabilistic model
- speech enhancement
- speaker adaptation
- noise model
- feature extraction
- speaker independent
- n gram
- similarity measure