Current Proceedings on Technology
Yazarlar: Surinder Dhanjal, Satvinder Singh Bhatia
Konular:-
Anahtar Kelimeler:Punjabi speech processing,Linear prediction,Lpc,Speech analysis and synthesis,Gurmukhi script,Malwai dialect
Özet: There has not yet been any work reported in the literature where the Punjabi speech has been analyzed/synthesized using the linear predictive coding (LPC). This work, therefore, concentrates on the linear prediction analysis and synthesis of the Punjabi speech. In the LP model of speech production, the current speech sample is predicted using the linear combination of the past p speech samples. Each of the four control parameters (LP coefficients, voiced/unvoiced decision, pitch period for voiced sounds, gain) can be determined directly from the speech waveform using more than one method, thereby making each analysis/synthesis project substantially different from every other project. Modern software tools (e.g., MATLAB and Praat), and methods (e.g., autocorrelation method) have been used in this work because of their graphics capabilities, efficiency, stability, simplicity and accuracy. Informal perceptual listening tests and spectral characteristics confirm that LPC technique is capable of generating high quality synthesized Punjabi speech.