A method of audio encoding and decoding for multiple audio objects

Abstract

The invention discloses a multi-audio-object-oriented audio coding and decoding method. During the coding process, firstly, multiple input audio object signals are converted from a time domain to a frequency domain, and the frequency domain signals of audio objects are subjected to noise reduction and combination to obtain a single-channel mixed signal. Secondly, parameters are extracted accordingto each audio object signal and the mixed signal. Thirdly, a large-size parameter matrix is decomposed into a small-size coefficient matrix and a base matrix by performing the matrix decomposition method. Finally, the mixed signal is synthesized with the coefficient matrix and the base matrix to obtain a code stream. During the decoding process, firstly, the code stream is decomposed, and a parameter matrix is obtained through reconstructing the coefficient matrix and the base matrix. Secondly, an audio object signal is reconstructed according to the parameter matrix and the mixed signal. According to the invention, the resolution of the parameter frequency domain is improved, and the code rate required by parameter coding is reduced by means of the matrix decomposition method. As a result, the audio object signal can be reconstructed at a low code rate and a high quality at the same time.

Description

technical field [0001] The invention belongs to the technical field of digital audio signal processing, and relates to an audio object encoding and decoding method, in particular to a multi-audio object compression encoding and decoding method with high-frequency domain resolution. Background technique [0002] Audio coding technology has made great progress, from traditional mono and stereo to more immersive multi-channel audio coding technology, such as MPEG spatial audio coding, speaker array and wave field synthesis. And based on multi-channel audio coding technology, the coding and reconstruction of 3D audio scene is realized, and it has received wide attention. However, the traditional audio scene coding technology based on channel coding has low flexibility and cannot meet the needs of users for personalized reconstruction of audio scenes and the accuracy requirements for single object trajectories. Therefore, in order to ensure the flexibility and accuracy of audio ...

AI Technical Summary

Problems solved by technology

[0004] Although the existing SAOC method based on a small number of sub-band extraction parameters can simultaneously encode multiple audio objects at a low bit rate, the object signals reconstructed by SAOC will be confused, resulting in poor sound quality and subjective listening experience

This is because the same subband shares a parameter so that the frequency components of the audio object signal cannot be accurately reconstructed, and the energy of multiple audio objects may be retained at certain frequency components, resulting in confusion

By increasing the number of sub-bands, the confusion between object signals can be effectively alleviated and the sound quality of the reconstructed audio object signal can be improved. However, simply increasing the number of sub-bands will increase the coding rate of audio objects

Therefore, the existing methods cannot guarantee the dual characteristics of low bit rate and high sound quality of audio object coding at the same time.

Images