CRL 707 Human and Machine Speech Communication
Introduction;(Human-machine speech communications aspects; speech chain,digital representations of speech;
intensity level of sound). Speech production (anatomy and physiology of speech organs; articulatory phonetics; acoustic phonetics; phonetics transcription; universal speech production model.) speech signal analysis (Time domain methods; Frequency domain methods; Pitch estimation spectrogram analysis; Cepstrum analysis;). Linear prediction coding (Least squares autocorrelation and covariance methods; Line spectral frequencies). Psychoacoustics and auditory perception (Hearing; critical bands; phenomena of masking; Mel scale; perceptually important features of speech; prosodic features).
Speech signal coding (Speech coder attributes; Coding rates; PCM; ADPCM; CELP; Harmonic coding of speech; Coding standards). Evaluation of speech quality(Dependancies of quality; Objective and subjective quality evaluation measures; Objective evaluation of subjective quality).
Speech synthesis (Limited and unrestricted text to speech synthesis; Articulatory synthesis; Concatenative synthesis; Incorporation of prosody). Automatic Speech recognition (Pattern recognition approach; Dynamic time warping; Feature extraction; HMM; Language models).
Speaker recognition(Verification vs. recognition, recognition techniques; features that distinguish speakers).
Back to Courses