A VERY-LOW COMPLEXITY VSELP SPEECH CODER USING REGULAR PULSE BASIS VECTORS

This paper describes a new Vector Sum Excited Linear Prediction (VSELP ) coder with very low complexity. The method, called regular pulse VSE LP (RP-VSELP), is based on regular pulse basis vectors with mutually o rthonormal property. In this Approach, a very efficient vector-sum cod ebook is constructed from a set of mutually orthonormal regular pulse bassis vectors and enables us to simplify the codebook search without additional degradation of synthesized speech compared with that of the conventional VSELP The regular pulse basis vectors are explicitly ort honormalized by means of the Gram-Schmidt procedure. To enhance the sp eech quality of the RP-VSELP speech coder, perceptually weighted disto rtion measure between the input and the synthesized speech is utilized in an iterative closed-loop training process of the regular pulse bas is vectors. It is shown that speech quality is improved by the trainin g process. Experimental results demonstrate that the proposed method p roduces the synthesized speech quality comparable to that of the VSELP scheme at the bit-rate of 4.8 Kbps.