APPARATUS AND METHOD FOR QUANTIZING SPEECH SIGNAL

摘要

PURPOSE: A voice signal quantization device and a method thereof for a voice signal according to a degree of the inputted voice signal are provided to supply the high coding efficiency about the voice signal by converting an index to a bit stream. CONSTITUTION: A KLT(Karhunen-Loeve Transform) unit changes an inputted voice signal into A KLT domain. A scalar quantization unit(310) quantizes the voice signal transformed to the KLT domain and generates the frame in which the index corresponding to the voice signal. The scalar quantization changes the voice signal into the code vector. An arithmetic coding unit(330) changes the code vector into a codeword consisting of the bit string.