Model-Based Speech Signal Coding Using Optimized Temporal Decomposition for Storage and Broadcasting Applications
In voice coding applications where there is no constraint on the encoding delay, segment coding techniques can be used to achieve a reduction in data rate. For low data rate linear predictive coding schemes, increasing the encoding delay allows one to exploit any long term temporal stationarities on an interframe basis, thus reducing the transmission bandwidth or storage needs of the speech signal. Transform coding has previously been applied in low data rate speech coding to exploit both the interframe and the intraframe correlation [ 1]. This paper investigates the potential of an adaptive transformation scheme for a segmented parametric speech representation. The problem of transform quantization is formulated and a solution methodology was proposed. The potential benefit of the use of the proposed adaptive transformation scheme is discussed in the context of segmented LSPs.