EC354
Speech Processing (3-0-0) 3
Nature of speech signal- Speech production mechanism. Classification of speech sounds. Digital processing of speech signals. Significance of short-time analysis.
Linear predictive coding of speech- linear prediction problem in time domain, normal equations, Relationship of linear prediction to autocorrelation and spectral domains.
Time domain and frequency domain methods for speech processing- methods for extracting the time-domain parameters. Zero crossings. Auto correlation function. pitch estimation. Short - time Fourier analysis. Filter bank analysis. Format extraction and pitch extraction. Analysis-synthesis systems.
Homomorphic speech analysis- Cepstral analysis, formant and pitch estimation. Applications to speech recognition and speaker identification. Basic pattern recognition methods, codebooks, HMM’s.
Speech codec standards and applications- Standards for low bit rate vocoders; Vocoder attributes. Encoders and decoders of G723.1, G726, G727, G728, G729 standard vocoders. Basics of voice over IP. Voice quality measures in IP networks.
References
T.F.Quatieri: Discrete-time Speech Signal Processing, Prentice-Hall, PTR, 2001.L.Hanza etal, Voice Compression and Communications, Wiley/IEEE , 2001.
L.R.Rabiner AND R.W.Schafer: Digital processing of speech signals, Prentice Hall,1978.
Hersent etal, IP Telephony, Pearson, 2000.