Speech coding and decoding system

摘要

A CELP type speech coding system which is provided with an arithmetic processing unit (31) which transforms a perceptual weighted input speech signal vector AX to a vector AAX, a sparse adaptive codebook (1) which stores a plurality of pitch prediction residual vectors P sparsed by a sparse unit, a multiplying unit (32) which multiplies the successively read out vectors P and the outcome AAX from the arithmetic processing unit (31), a filter operation unit (33) which performs a filter operation on the vectors P and an evaluation unit (34) which finds the optimum vector P based on the output from the filter operation unit (33), so as to enable reduction of the amount of arithmetic operations.