Decomposition of voiced speech into all-pole filter and sinusoidal excitation | IEEE Conference Publication | IEEE Xplore