Method for enhancing excitation signal naturalism based on judgment and processing of transition frames

Abstract

The invention provides a method for improving the fidelity of an actuating signal on the basis of transition frame judgment and treatment, which belongs to the technical field of the compressed encoding of low-rate sounds. In the case that the ratio of the average energy of 60 sampling points before the current frame to the average energy of the 60 sampling points after the current frame is less than 1 / 32, then the current frame is judged to be a transition frame and expressed by the redundancy mode of sub-band surd and sonant vectors; a decoding end carries out de-quantification to parameters and the sub-band surd and sonant vectors obtained by decoding are utilized to judge if the current frame is the transition frame, if no, then judge if the current frame is a sonant frame and if the previous frame is a surd frame; if yes, the parameters of the current frame at the decoding end is not interpolated with the parameters of the previous frame when the actuating signal is synthesized. The method can improve the fidelity of a synthesized speech and is suitable for an SELP 2.4kbps vocoder.

Description

technical field [0001] The invention belongs to the technical field of speech coding, and is particularly aimed at SELP 2.4kbps vocoder technology. Background technique [0002] Speech coding is widely used in communication systems, voice storage and playback systems, and consumer products with voice functions. In recent years, the International Telecommunication Union (ITU), some regional organizations and some countries have successively formulated a series of speech compression coding standards, and obtained satisfactory speech quality at the coding rate of 2.4kb / s to 16kb / s. At present, research at home and abroad is mainly focused on high-quality voice compression coding at a rate below 2.4kb / s, which is mainly used for wireless communication, secure communication, and large-capacity voice storage and playback. The synthesis of excitation signals in low-rate speech coding is very important. SELP vocoder uses mixed excitation signals, and uses pitch period parameters, e...

AI Technical Summary

Problems solved by technology

[0013] like figure 1 As shown, in the quantization of unvoiced and voiced sound decision parameters in sub-bands, the original technology adopts a simple 1-bit quantization method of 5-bit quantization for each sub-band unvoiced and voiced sound decision parameter, which will cause a certain degree of redundancy in the coded stream. underutilization of bits

Images