CN-121999790-A - Audio self-adaptive noise reduction method, device, system, electronic equipment and storage medium

CN121999790ACN 121999790 ACN121999790 ACN 121999790ACN-121999790-A

Abstract

The disclosure relates to an audio self-adaptive noise reduction method, an audio self-adaptive noise reduction device, an audio self-adaptive noise reduction system, electronic equipment and a storage medium, wherein the method comprises the steps of estimating a first signal-to-noise ratio of an audio input signal to obtain first signal-to-noise ratio information; the method comprises the steps of estimating a second signal-to-noise ratio of an audio output signal to obtain second signal-to-noise ratio information, selecting one noise reduction operation pipeline from at least two preset noise reduction operation pipelines according to the first signal-to-noise ratio information and the second signal-to-noise ratio information, and performing noise reduction operation on the audio input signal based on the selected noise reduction operation pipeline to obtain the audio output signal. The method and the device realize the self-adaptive adjustment of the noise reduction mode according to the change of the signal-to-noise ratio of the audio input signal, and are beneficial to obtaining the expected noise reduction effect on the audio input signal before and after the conversion between different scenes.

Inventors

YUAN BIN
LING MING
AI GUO
YANG ZUOXING

Assignees

深圳比特微电子科技有限公司

Dates

Publication Date: 20260508
Application Date: 20241107

Claims (11)

1. An audio adaptive noise reduction method, comprising: acquiring an audio input signal to be processed; Performing first signal-to-noise ratio estimation on the audio input signal to obtain first signal-to-noise ratio information; Performing second signal-to-noise ratio estimation on an audio output signal to obtain second signal-to-noise ratio information, wherein the audio output signal is obtained by performing noise reduction operation on the audio input signal; Selecting one noise reduction operation pipeline from at least two preset noise reduction operation pipelines according to the first signal-to-noise ratio information and the second signal-to-noise ratio information, and And performing noise reduction operation on the audio input signal based on the selected noise reduction operation pipeline to obtain the audio output signal.
2. The audio adaptive noise reduction method according to claim 1, wherein: the at least two noise reduction operation pipelines comprise a first noise reduction operation sub-pipeline and a second noise reduction operation sub-pipeline; Wherein the first noise reduction operation sub-pipeline includes an NR process and an AGC process performed sequentially on the audio input signal; The second noise reduction operation sub-pipeline includes AGC processing and NR processing performed sequentially on the audio input signal.
3. The method of adaptive noise reduction according to claim 2, wherein selecting one noise reduction operation pipeline from at least two noise reduction operation pipelines preset according to the first signal-to-noise ratio information and the second signal-to-noise ratio information comprises: Selecting the first noise reduction operation sub-pipeline under the condition that the first signal-to-noise ratio information is larger than or equal to a preset first signal-to-noise ratio threshold value; Selecting the second noise reduction operation sub-pipeline under the condition that the first signal-to-noise ratio information is less than or equal to a preset second signal-to-noise ratio threshold value, or And under the condition that the first signal-to-noise ratio information is between the first signal-to-noise ratio threshold value and the second signal-to-noise ratio threshold value, selecting one of the first noise reduction operation sub-pipeline and the second noise reduction operation sub-pipeline according to the second signal-to-noise ratio information.
4. The method of audio adaptive noise reduction according to claim 3, wherein said selecting one of the first noise reduction operation sub-pipeline and the second noise reduction operation sub-pipeline based on the second signal-to-noise ratio information if the first signal-to-noise ratio information is between the first signal-to-noise ratio threshold and the second signal-to-noise ratio threshold comprises: Obtaining the change rate of the second signal-to-noise ratio information according to a plurality of pieces of the second signal-to-noise ratio information obtained in a preset time window width range; selecting the current noise reduction operation pipeline for performing noise reduction operation on the audio input signal in the first noise reduction operation sub-pipeline and the second noise reduction operation sub-pipeline under the condition that the change rate of the second signal-to-noise ratio information is smaller than or equal to a preset change rate threshold value, or And selecting another noise reduction operation pipeline except the current noise reduction operation pipeline in the first noise reduction operation sub-pipeline and the second noise reduction operation sub-pipeline under the condition that the change rate of the second signal-to-noise ratio information is larger than or equal to a preset change rate threshold.
5. The method of adaptive noise reduction according to claim 1, wherein the performing a first signal-to-noise ratio estimation on the audio input signal to obtain first signal-to-noise ratio information includes: framing the audio input signal to obtain an audio frame of the input signal; Extracting time domain features and frequency domain features of the input signal audio frame; obtaining the voice existence probability and the noise existence probability in the audio input signal according to the time domain characteristics and the frequency domain characteristics of the audio frame of the input signal; And obtaining the first signal-to-noise ratio information according to the voice existence probability and the noise existence probability in the audio input signal.
6. The method of adaptive noise reduction according to claim 1, wherein the performing a second signal-to-noise ratio estimation on the audio output signal to obtain second signal-to-noise ratio information includes: framing the audio output signal to obtain an audio frame of the output signal; extracting time domain features and frequency domain features of the output signal audio frame; Obtaining the voice existence probability and the noise existence probability in the audio output signal according to the time domain characteristics and the frequency domain characteristics of the audio frame of the output signal; And obtaining the second signal-to-noise ratio information according to the voice existence probability and the noise existence probability in the audio output signal.
7. The audio adaptive noise reduction method according to claim 1, characterized in that after the noise reduction operation is performed on the audio input signal based on the selected noise reduction operation pipeline, and before the audio output signal is obtained, the audio adaptive noise reduction method further comprises: performing voice smooth transition processing on the audio signal obtained after the noise reduction operation is performed; the audio output signal is an audio signal obtained after the voice smooth transition processing.
8. An audio adaptive noise reduction device, comprising: an audio acquisition module configured to perform acquisition of an audio input signal to be processed; The first signal-to-noise ratio estimation module is configured to perform first signal-to-noise ratio estimation on the audio input signal to obtain first signal-to-noise ratio information; The second signal-to-noise ratio estimation module is configured to perform second signal-to-noise ratio estimation on an audio output signal to obtain second signal-to-noise ratio information, wherein the audio output signal is obtained by performing noise reduction operation on the audio input signal; A noise reduction operation pipeline selection module configured to perform selection of one noise reduction operation pipeline from at least two noise reduction operation pipelines preset according to the first signal-to-noise ratio information and the second signal-to-noise ratio information, and And the noise reduction execution module is configured to execute noise reduction operation on the audio input signal based on the selected noise reduction operation pipeline to obtain the audio output signal.
9. An audio adaptive noise reduction system, comprising: The audio acquisition module is used for acquiring an audio input signal to be processed; the first signal-to-noise ratio estimation module is used for carrying out first signal-to-noise ratio estimation on the audio input signal to obtain first signal-to-noise ratio information; The noise reduction operation pipeline module comprises at least two preset noise reduction operation pipelines, and the at least two noise reduction operation pipelines are used for performing noise reduction processing on the audio input signal to obtain an audio output signal; A noise reduction operation scheduling module, configured to select one noise reduction operation pipeline from the at least two noise reduction operation pipelines according to the first signal-to-noise ratio information and the second signal-to-noise ratio information related to the audio output signal, and input the audio input signal into the selected noise reduction operation pipeline; The second signal-to-noise ratio estimation module is used for carrying out second signal-to-noise ratio estimation on the audio output signal to obtain second signal-to-noise ratio information; And the audio output module is used for outputting the audio output signal.
10. An electronic device, comprising: A processor; A memory for storing executable instructions of the processor; wherein the processor is configured to execute the executable instructions to implement the audio adaptive noise reduction method of any of claims 1 to 7.
11. A computer readable storage medium, characterized in that at least one instruction in the computer readable storage medium, when executed by a processor of an electronic device, enables the electronic device to implement the audio adaptive noise reduction method of any one of claims 1 to 7.

Description

Audio self-adaptive noise reduction method, device, system, electronic equipment and storage medium Technical Field The present disclosure relates to the field of audio processing, and in particular, to an audio adaptive noise reduction method, apparatus, system, electronic device, and storage medium. Background In two different environments, indoor and outdoor, in order to improve the quality of the collected voice, the image-voice collection system (e.g., camera system) may perform noise reduction processing on the collected audio in different manners. If the noise reduction processing is performed on the audio collected in the outdoor environment by adopting the noise reduction processing mode for the indoor environment or the noise reduction processing is performed on the audio collected in the indoor environment by adopting the noise reduction processing mode for the outdoor environment, the problem that the noise reduction effect is poor and the expected effect cannot be achieved can occur. Disclosure of Invention In view of this, the present disclosure provides an audio adaptive noise reduction method, apparatus, system, electronic device and storage medium, which implement adaptive adjustment of noise reduction according to the change of the signal-to-noise ratio of an audio input signal, so as to help achieve the expected noise reduction effect on the audio input signal before and after the transition between different scenes. The technical scheme of the present disclosure is realized as follows: according to an aspect of the embodiments of the present disclosure, there is provided an audio adaptive noise reduction method, including: acquiring an audio input signal to be processed; Performing first signal-to-noise ratio estimation on the audio input signal to obtain first signal-to-noise ratio information; Performing second signal-to-noise ratio estimation on an audio output signal to obtain second signal-to-noise ratio information, wherein the audio output signal is obtained by performing noise reduction operation on the audio input signal; Selecting one noise reduction operation pipeline from at least two preset noise reduction operation pipelines according to the first signal-to-noise ratio information and the second signal-to-noise ratio information, and And performing noise reduction operation on the audio input signal based on the selected noise reduction operation pipeline to obtain the audio output signal. In one possible implementation, the at least two noise reduction operation pipelines include a first noise reduction operation sub-pipeline and a second noise reduction operation sub-pipeline; Wherein the first noise reduction operation sub-pipeline includes an NR process and an AGC process performed sequentially on the audio input signal; The second noise reduction operation sub-pipeline includes AGC processing and NR processing performed sequentially on the audio input signal. In one possible implementation manner, the selecting one noise reduction operation pipeline from at least two preset noise reduction operation pipelines according to the first signal-to-noise ratio information and the second signal-to-noise ratio information includes: Selecting the first noise reduction operation sub-pipeline under the condition that the first signal-to-noise ratio information is larger than or equal to a preset first signal-to-noise ratio threshold value; Selecting the second noise reduction operation sub-pipeline under the condition that the first signal-to-noise ratio information is less than or equal to a preset second signal-to-noise ratio threshold value, or And under the condition that the first signal-to-noise ratio information is between the first signal-to-noise ratio threshold value and the second signal-to-noise ratio threshold value, selecting one of the first noise reduction operation sub-pipeline and the second noise reduction operation sub-pipeline according to the second signal-to-noise ratio information. In one possible implementation manner, the selecting one of the first noise reduction operation sub-pipeline and the second noise reduction operation sub-pipeline according to the second signal to noise ratio information under the condition that the first signal to noise ratio information is between the first signal to noise ratio threshold and the second signal to noise ratio threshold includes: Obtaining the change rate of the second signal-to-noise ratio information according to a plurality of pieces of the second signal-to-noise ratio information obtained in a preset time window width range; selecting the current noise reduction operation pipeline for performing noise reduction operation on the audio input signal in the first noise reduction operation sub-pipeline and the second noise reduction operation sub-pipeline under the condition that the change rate of the second signal-to-noise ratio information is smaller than or equal to a preset change rate threshold value, or And