Title :
Prediction in speech coding: the modification of the coding of LPC parameters and nonlinear estimation technique by using ANN
Author_Institution :
Dept. of Electr. Eng., Nanjing Univ. of Posts & Telecommun., Nanjing, China
Abstract :
Linear and non-linear prediction techniques of speech signal and the different performance of LPC algorithms are discussed. A new coding plan has been suggested to raise the quantification SNR by about 2 dB by coding the difference of the log-area ratio (LAR) of adjacent frames instead of the LAR. Then, the stable time delay ANN is used to carry out NLP (non-linear prediction) tasks. Experiment results show that only one NLP system is enough to deprive all short-term and long-term correlation from the speech samples and the energy from the output of the NLP system is much smaller than that obtained with the LP system
Keywords :
delays; linear predictive coding; neural nets; parameter estimation; signal sampling; speech coding; speech processing; stability; LPC algorithms; LPC parameters; NLP system; energy; experiment results; linear prediction; log-area ratio; nonlinear estimation; nonlinear prediction; speech coding; speech samples correlation; speech signal; stable time delay ANN; Delay effects; Linear predictive coding; Mobile communication; Nonlinear filters; Parameter estimation; Signal processing algorithms; Speech coding; Speech enhancement; Telephony; White noise;
Conference_Titel :
Signal Processing, 1996., 3rd International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-2912-0
DOI :
10.1109/ICSIGP.1996.567357