A generative confrontation network training method and audio encoding and decoding method for frequency band expansion

Abstract

The invention discloses a generative confrontation network training method and an audio encoding and decoding method oriented to frequency band expansion. The training method of the generative confrontation network of the present invention is as follows: detect the transient signal of the audio signal; then perform MDCT transformation on it according to the detection result, and use the obtained spectrum as real data; divide the spectrum into bands, and calculate the high and low frequency spectrum Energy envelope ratio, and then quantize and inverse quantize the energy envelope ratio of the high and low frequency spectrum; input the low frequency spectrum obtained by band division into the generation network GAN to generate high frequency spectrum; use the dequantized high frequency energy envelope to correct the generated The high-frequency spectrum is obtained to obtain the final generated high-frequency spectrum; the final generated high-frequency spectrum and the low-frequency spectrum obtained by banding are synthesized into a full-band generated spectrum, and the full-band generated spectrum is used as false data; the real data, Fake data is used as the input of the discriminative network D to train the generative adversarial network. The network trained by the invention is easy to converge.

Description

technical field [0001] The invention belongs to the field of audio coding and decoding, and relates to a frequency band expansion method, in particular to a frequency band expansion-oriented generative confrontation network training method, an audio coding method, and a decoding method. Background technique [0002] Audio codec technology, also known as audio compression technology, compresses and encodes audio files to reduce the file bit rate, making the results easy to record, store, and transmit, and has a wide range of uses. When the target bit rate is low, the traditional mono audio codec technology will discard high-frequency information to ensure low-frequency compression effect, but due to the lack of high-frequency information, the sound of the codec result will cause hollowness, dullness and other discomfort a feeling of. In order to improve the codec quality, the decoding result of the single-channel core encoder is usually band-extended. Such methods are colle...

AI Technical Summary

Problems solved by technology

Aiming at the shortcomings of generative adversarial networks that are not easy to converge and the particularity of the sound signal frequency band expansion task, the traditional generative adversarial network is improved by introducing real low-frequency information and high-frequency envelopes, and a complete single-channel network is built on this basis codec system

Images