Title :
Interpolation of LSP coefficients using recurrent neural networks
Author_Institution :
Fac. of Eng., Tohoku Univ., Sendai, Japan
fDate :
8/1/1996 12:00:00 AM
Abstract :
To reduce the bit rate in very low bit rate speech coding, recurrent neural networks (RNNs) were applied to interpolate LSP coefficients, and it was possible to increase the duration of interpolation to ~100 ms. Simulation of coding was carried out, and the proposed method outperformed the other interpolation methods below a bit rate of 300 bit/s
Keywords :
interpolation; prediction theory; recurrent neural nets; speech coding; 100 ms; LSP coefficients; bit rate; coding simulation; interpolation duration; line spectrum pair; recurrent neural networks; speech coding;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19961009