Audio processing method and device, electronic device and computer readable storage medium

Abstract

The present application discloses an audio processing method and device, an electronic device, and a computer-readable storage medium. The method includes: obtaining a first audio to be tested and a target audio processing algorithm; processing the first audio to be tested using the target audio processing algorithm to obtain first result data, the first result data includes a first result audio, and the first result audio is the audio obtained by processing the first audio to be tested using the target audio processing algorithm; according to the speech in the first result audio, determining the processing effect of the target audio processing algorithm on the speech, and obtaining a first evaluation result; according to the non-speech data in the first result data, determining the processing effect of the target audio processing algorithm on the audio, and obtaining a second evaluation result; according to the first evaluation result and the second evaluation result, determining the processing effect of the target audio processing algorithm on the audio. The processing effect of the audio processing algorithm can be evaluated by this method.

Description

Technical Field

[0001] The present application relates to the field of audio processing technology, and in particular to an audio processing method and device, an electronic device, and a computer-readable storage medium. Background Art

[0002] In practical applications, audio processing algorithms are usually required to process audio to meet demand. Therefore, how to evaluate the processing effect of audio processing algorithms is of great significance. Summary of the invention

[0003] The present application provides an audio processing method and device, an electronic device, and a computer-readable storage medium.

[0004] In a first aspect, an audio processing method is provided, the method comprising:

[0005] Obtaining a first audio to be tested and a target audio processing algorithm;

[0006] Processing the first audio to be tested by using a target audio processing algorithm to obtain first result data, where the first result data includes a first result audio, where the first result audio is audio obtained by processing the first audio to be tested by using the target audio processing algorithm;

[0007] Determine, according to the speech in the first result audio, the processing effect of the target audio processing algorithm on the speech, and obtain a first evaluation result;

[0008] Determine the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data, and obtain a second evaluation result;

[0009] The processing effect of the target audio processing algorithm on the audio is determined according to the first evaluation result and the second evaluation result.

[0010] In combination with any embodiment of the present application, determining the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data to obtain the second evaluation result includes:

[0011] By detecting the acoustic features of the first result audio, the acoustic features of the first result audio are obtained;

[0012] The second evaluation result is obtained according to the acoustic feature.

[0013] In combination with any embodiment of the present application, the acoustic feature includes one or more of sound pressure level and loudness;

[0014] The obtaining the second evaluation result according to the acoustic feature includes:

[0015] In the case where the acoustic feature includes a sound pressure level, determining a sound pressure level dynamic range according to a maximum value of the sound pressure level and a minimum value of the sound pressure level, the sound pressure level dynamic range being a dynamic range of the sound pressure level of the first result audio;

[0016] Obtaining the second evaluation result according to the sound pressure level dynamic range;

[0017] In the case where the acoustic feature includes loudness, determining a loudness dynamic range according to a maximum value of the loudness and a minimum value of the loudness, wherein the loudness dynamic range is a dynamic range of the loudness of the first result audio;

[0018] The second evaluation result is obtained according to the loudness dynamic range.

[0019] In combination with any embodiment of the present application, the non-speech data includes one or more of a target duration and a target calculation amount, the target duration represents the time consumed by the target audio processing algorithm to obtain the first result audio, and the target calculation amount is the calculation amount generated by obtaining the first result audio through the target audio processing algorithm;

[0020] The step of determining the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data to obtain a second evaluation result includes:

[0021] In a case where the non-speech data includes the time, determining the second evaluation result according to the time;

[0022] In a case where the non-speech data includes the calculation amount, the second evaluation result is determined according to the calculation amount.

[0023] In combination with any embodiment of the present application, determining the processing effect of the target audio processing algorithm on the speech according to the speech in the first result audio to obtain the first evaluation result includes:

[0024] Determining the quality of the speech in the first result audio;

[0025] Determine, based on the quality, a processing effect of the target audio processing algorithm on the speech, and obtain the first evaluation result;

[0026] Alternatively, determining speech intelligibility of speech in the first result audio;

[0027] According to the speech intelligibility, the processing effect of the target audio processing algorithm on the speech is determined to obtain the first evaluation result.

[0028] In combination with any embodiment of the present application, before determining the processing effect of the target audio processing algorithm on the speech according to the speech in the first result audio and obtaining the first evaluation result, the method further includes:

[0029] Acquire voice audio, where the voice audio is audio including voice, and the first audio to be tested is audio obtained based on the voice audio;

[0030] The step of determining, based on the speech in the first result audio, a processing effect of the target audio processing algorithm on the speech to obtain a first evaluation result includes:

[0031] In a case where the energy of the speech audio is less than the energy of the first result audio, determining the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, the noise in the first audio to be tested remains in the first result audio;

[0032] When the energy of the speech audio is greater than the energy of the first result audio, determining the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, speech distortion exists in the first audio to be tested.

[0033] In combination with any implementation manner of the present application, the obtaining of the first audio to be tested includes:

[0034] Acquire music audio, where the sound in the music audio is music;

[0035] By mixing the voice audio, the music audio and the noise, an audio satisfying a preset requirement is obtained as the first audio to be tested.

[0036] In combination with any implementation manner of the present application, determining the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result includes:

[0037] Obtaining a first reference range and a second reference range, wherein the first reference range is a normal range of the first evaluation result, and the second reference range is a normal range of the second evaluation result;

[0038] When the first evaluation result is outside the first reference range, or the second evaluation result is outside the second reference range, determining the processing effect of the target audio processing algorithm on the audio includes that the target audio processing algorithm processes the target class audio abnormally, and the target class audio is audio that meets the preset requirements.

[0039] In combination with any implementation manner of the present application, obtaining the first reference range includes:

[0040] Obtain at least one third evaluation result, wherein the third evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the speech according to the speech in the second result audio, the second result audio is the audio obtained by processing the second audio to be tested by the target audio processing algorithm, and the first audio to be tested is different from the second audio to be tested; the third evaluation result may be a result obtained by determining the first audio processing algorithm to be tested according to the speech in the third result audio, and the third result audio is the audio obtained by processing the audio by the first audio processing algorithm to be tested;

[0041] Arrange the first evaluation result and the at least one third evaluation result in order from largest to smallest to obtain an arrangement result;

[0042] The first reference range is determined according to a first preset quantile in the arrangement result and a second preset quantile in the arrangement result.

[0043] In combination with any implementation manner of the present application, determining the first reference range according to the first preset quantile in the arrangement result and the second preset quantile in the arrangement result includes:

[0044] Determining a first reference value corresponding to the first preset quantile from the arrangement result;

[0045] Determining a second reference value corresponding to the second preset quantile from the arrangement result;

[0046] The first reference range is determined according to the first reference value and the second reference value.

[0047] In combination with any implementation manner of the present application, determining the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result includes:

[0048] Obtain a fourth evaluation result and a fifth evaluation result, wherein the fourth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the speech based on the speech in the fourth result audio, and the fourth result audio is obtained by processing the first audio to be evaluated by using the second audio processing algorithm to be evaluated, and the second audio processing algorithm to be evaluated is different from the target audio processing algorithm; the fifth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the audio based on the non-speech data in the second result data, and the second result data is obtained by processing the first audio to be tested by using the second audio processing algorithm to be evaluated;

[0049] Aggregating the first evaluation result and the fourth evaluation result to obtain a first aggregated result;

[0050] Aggregating the second evaluation result and the fifth evaluation result to obtain a second aggregated result;

[0051] According to the first aggregation result and the second aggregation result, a difference between a processing effect of the target audio processing algorithm and a processing effect of the second audio processing algorithm to be evaluated is determined.

[0052] In combination with any implementation manner of the present application, determining the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result includes:

[0053] Obtain a sixth evaluation result and a seventh evaluation result, wherein the sixth evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the speech based on the speech in the fifth result audio, and the fifth result audio is obtained by processing the second audio to be tested using the target audio processing algorithm, and the second audio to be tested is different from the first audio to be tested; the seventh evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the audio based on the non-speech data in the third result data, and the third result data is obtained by processing the second audio to be tested using the target audio processing algorithm;

[0054] Aggregating the first evaluation result and the sixth evaluation result to obtain a third aggregated result;

[0055] Aggregating the second evaluation result and the seventh evaluation result to obtain a fourth aggregated result;

[0056] The processing effect of the target audio processing algorithm on different audios is determined according to the third aggregation result and the fourth aggregation result.

[0057] In combination with any implementation manner of the present application, determining the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result includes:

[0058] When the number of the first evaluation results is greater than 1, weighted fusion is performed on the first evaluation results to obtain an eighth evaluation result;

[0059] Determining, according to the eighth evaluation result and the second evaluation result, a processing effect of the target audio processing algorithm on the audio;

[0060] Alternatively, when the number of the second evaluation results is greater than 1, weighted fusion is performed on the second evaluation results to obtain a ninth evaluation result;

[0061] The processing effect of the target audio processing algorithm on the audio is determined according to the first evaluation result and the eighth evaluation result.

[0062] In combination with any implementation of the present application, obtaining a target audio processing algorithm includes:

[0063] Acquire target configuration parameters of the target audio processing algorithm, where the target configuration parameters include at least one candidate audio processing algorithm required to obtain the target audio processing algorithm, a coupling relationship between the at least one candidate audio processing algorithm, and at least one candidate configuration parameter of the at least one candidate audio processing algorithm;

[0064] The at least one candidate audio processing algorithm is configured according to the at least one candidate configuration parameter and the coupling relationship to obtain the target audio processing algorithm.

[0065] In a second aspect, an audio processing device is provided, the audio processing device comprising:

[0066] An acquisition unit, used to acquire a first audio to be tested and a target audio processing algorithm;

[0067] a processing unit, configured to process the first audio to be tested by using a target audio processing algorithm to obtain first result data, wherein the first result data includes a first result audio, and the first result audio is an audio obtained by processing the first audio to be tested by using the target audio processing algorithm;

[0068] A determination unit, configured to determine, based on the speech in the first result audio, a processing effect of the target audio processing algorithm on the speech, and obtain a first evaluation result;

[0069] The determining unit is used to determine the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data, so as to obtain a second evaluation result;

[0070] The determining unit is used to determine the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result.

[0071] In combination with any implementation manner of the present application, the determining unit is specifically configured to:

[0072] By detecting the acoustic features of the first result audio, the acoustic features of the first result audio are obtained;

[0073] The second evaluation result is obtained according to the acoustic feature.

[0074] In combination with any embodiment of the present application, the acoustic feature includes one or more of sound pressure level and loudness;

[0075] The determining unit is specifically configured to:

[0076] In the case where the acoustic feature includes a sound pressure level, determining a sound pressure level dynamic range according to a maximum value of the sound pressure level and a minimum value of the sound pressure level, the sound pressure level dynamic range being a dynamic range of the sound pressure level of the first result audio;

[0077] Obtaining the second evaluation result according to the sound pressure level dynamic range;

[0078] In the case where the acoustic feature includes loudness, determining a loudness dynamic range according to a maximum value of the loudness and a minimum value of the loudness, wherein the loudness dynamic range is a dynamic range of the loudness of the first result audio;

[0079] The second evaluation result is obtained according to the loudness dynamic range.

[0080] In combination with any embodiment of the present application, the non-speech data includes one or more of a target duration and a target calculation amount, the target duration represents the time consumed by the target audio processing algorithm to obtain the first result audio, and the target calculation amount is the calculation amount generated by obtaining the first result audio through the target audio processing algorithm;

[0081] The determining unit is specifically configured to:

[0082] In a case where the non-speech data includes the time, determining the second evaluation result according to the time;

[0083] In a case where the non-speech data includes the calculation amount, the second evaluation result is determined according to the calculation amount.

[0084] In combination with any implementation manner of the present application, the determining unit is specifically configured to:

[0085] Determining the quality of the speech in the first result audio;

[0086] Determine, based on the quality, a processing effect of the target audio processing algorithm on the speech, and obtain the first evaluation result;

[0087] Alternatively, determining speech intelligibility of speech in the first result audio;

[0088] According to the speech intelligibility, the processing effect of the target audio processing algorithm on the speech is determined to obtain the first evaluation result.

[0089] In combination with any embodiment of the present application, the acquisition unit is further used to acquire voice audio, where the voice audio is audio including voice, and the first audio to be tested is audio obtained based on the voice audio;

[0090] The determining unit is specifically configured to:

[0091] In a case where the energy of the speech audio is less than the energy of the first result audio, determining the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, the noise in the first audio to be tested remains in the first result audio;

[0092] When the energy of the speech audio is greater than the energy of the first result audio, determining the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, speech distortion exists in the first audio to be tested.

[0093] In combination with any implementation manner of the present application, the acquisition unit is specifically used to:

[0094] Acquire music audio, where the sound in the music audio is music;

[0095] By mixing the voice audio, the music audio and the noise, an audio satisfying a preset requirement is obtained as the first audio to be tested.

[0096] In combination with any implementation manner of the present application, the determining unit is specifically configured to:

[0097] Obtaining a first reference range and a second reference range, wherein the first reference range is a normal range of the first evaluation result, and the second reference range is a normal range of the second evaluation result;

[0098] When the first evaluation result is outside the first reference range, or the second evaluation result is outside the second reference range, determining the processing effect of the target audio processing algorithm on the audio includes that the target audio processing algorithm processes the target class audio abnormally, and the target class audio is audio that meets the preset requirements.

[0099] In combination with any implementation manner of the present application, the acquisition unit is specifically used to:

[0100] Obtain at least one third evaluation result, wherein the third evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the speech according to the speech in the second result audio, the second result audio is the audio obtained by processing the second audio to be tested by the target audio processing algorithm, and the first audio to be tested is different from the second audio to be tested; the third evaluation result may be a result obtained by determining the first audio processing algorithm to be tested according to the speech in the third result audio, and the third result audio is the audio obtained by processing the audio by the first audio processing algorithm to be tested;

[0101] Arrange the first evaluation result and the at least one third evaluation result in order from largest to smallest to obtain an arrangement result;

[0102] The first reference range is determined according to a first preset quantile in the arrangement result and a second preset quantile in the arrangement result.

[0103] In combination with any implementation manner of the present application, the determining unit is specifically configured to:

[0104] Determining a first reference value corresponding to the first preset quantile from the arrangement result;

[0105] Determining a second reference value corresponding to the second preset quantile from the arrangement result;

[0106] The first reference range is determined according to the first reference value and the second reference value.

[0107] In combination with any implementation manner of the present application, the determining unit is specifically configured to:

[0108] Obtain a fourth evaluation result and a fifth evaluation result, wherein the fourth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the speech based on the speech in the fourth result audio, and the fourth result audio is obtained by processing the first audio to be evaluated by using the second audio processing algorithm to be evaluated, and the second audio processing algorithm to be evaluated is different from the target audio processing algorithm; the fifth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the audio based on the non-speech data in the second result data, and the second result data is obtained by processing the first audio to be tested by using the second audio processing algorithm to be evaluated;

[0109] Aggregating the first evaluation result and the fourth evaluation result to obtain a first aggregated result;

[0110] Aggregating the second evaluation result and the fifth evaluation result to obtain a second aggregated result;

[0111] According to the first aggregation result and the second aggregation result, a difference between a processing effect of the target audio processing algorithm and a processing effect of the second audio processing algorithm to be evaluated is determined.

[0112] In combination with any implementation manner of the present application, the determining unit is specifically configured to:

[0113] Obtain a sixth evaluation result and a seventh evaluation result, wherein the sixth evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the speech based on the speech in the fifth result audio, and the fifth result audio is obtained by processing the second audio to be tested using the target audio processing algorithm, and the second audio to be tested is different from the first audio to be tested; the seventh evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the audio based on the non-speech data in the third result data, and the third result data is obtained by processing the second audio to be tested using the target audio processing algorithm;

[0114] Aggregating the first evaluation result and the sixth evaluation result to obtain a third aggregated result;

[0115] Aggregating the second evaluation result and the seventh evaluation result to obtain a fourth aggregated result;

[0116] The processing effect of the target audio processing algorithm on different audios is determined according to the third aggregation result and the fourth aggregation result.

[0117] In combination with any implementation manner of the present application, the determining unit is specifically configured to:

[0118] When the number of the first evaluation results is greater than 1, weighted fusion is performed on the first evaluation results to obtain an eighth evaluation result;

[0119] Determining, according to the eighth evaluation result and the second evaluation result, a processing effect of the target audio processing algorithm on the audio;

[0120] Alternatively, when the number of the second evaluation results is greater than 1, weighted fusion is performed on the second evaluation results to obtain a ninth evaluation result;

[0121] The processing effect of the target audio processing algorithm on the audio is determined according to the first evaluation result and the eighth evaluation result.

[0122] In combination with any implementation manner of the present application, the acquisition unit is specifically used to:

[0123] Acquire target configuration parameters of the target audio processing algorithm, where the target configuration parameters include at least one candidate audio processing algorithm required to obtain the target audio processing algorithm, a coupling relationship between the at least one candidate audio processing algorithm, and at least one candidate configuration parameter of the at least one candidate audio processing algorithm;

[0124] The at least one candidate audio processing algorithm is configured according to the at least one candidate configuration parameter and the coupling relationship to obtain the target audio processing algorithm.

[0125] In a third aspect, an electronic device is provided, comprising: a processor and a memory, the memory being used to store computer program code, the computer program code comprising computer instructions, and when the processor executes the computer instructions, the electronic device executes the method as described in the first aspect above and any possible implementation thereof.

[0126] In a fourth aspect, another electronic device is provided, comprising: a processor, a sending device, an input device, an output device and a memory, wherein the memory is used to store computer program code, and the computer program code includes computer instructions. When the processor executes the computer instructions, the electronic device executes the first aspect and any one of its embodiments.

[0127] In a fifth aspect, a computer-readable storage medium is provided, in which a computer program is stored. The computer program includes program instructions. When the program instructions are executed by a processor, the processor is caused to execute the first aspect and any one of its embodiments.

[0128] In a sixth aspect, a computer program product is provided, wherein the computer program product comprises a computer program or instructions, and when the computer program or instructions are run on a computer, the computer is enabled to execute the above-mentioned first aspect and any embodiment thereof.

[0129] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application.

[0130] In the present application, after acquiring the first audio to be tested and the target audio processing algorithm, the audio processing device processes the first audio to be tested using the target audio processing algorithm to obtain first result data, wherein the first result data includes the first result audio obtained by processing the first audio to be tested using the target audio processing algorithm. The audio processing device then determines the processing effect of the target audio processing algorithm on the speech based on the speech in the first result audio to obtain a first evaluation result, and determines the processing effect of the target audio processing algorithm on the audio based on the non-speech data in the first result data to obtain a second evaluation result. Finally, based on the first evaluation result and the second evaluation result, the processing effect of the target audio processing algorithm on the audio is determined, and the processing effect of the target audio processing algorithm on the audio can be determined based on the speech in the first result data and the non-speech data in the first result data, thereby improving the accuracy of the evaluation of the processing effect of the target audio processing algorithm on the audio. BRIEF DESCRIPTION OF THE DRAWINGS

[0131] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background technology, the drawings required for use in the embodiments of the present application or the background technology will be described below.

[0132] The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and are used together with the specification to illustrate the technical solution of the present application.

[0133] Figure 1 A flowchart of an audio processing method provided in an embodiment of the present application;

[0134] Figure 2 A flowchart of another audio processing method provided in an embodiment of the present application;

[0135] Figure 3 A flowchart of another audio processing method provided in an embodiment of the present application;

[0136] Figure 4 A schematic diagram of the structure of an audio processing device provided in an embodiment of the present application;

[0137] Figure 5 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0138] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0139] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0140] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0141] The execution subject of the embodiment of the present application is an audio processing device, wherein the audio processing device can be any electronic device that can execute the technical solution disclosed in the embodiment of the method of the present application. Optionally, the audio processing device can be one of the following: a computer, a server.

[0142] It should be understood that the method embodiment of the present application can also be implemented by a processor executing a computer program code. The following describes the embodiment of the present application in conjunction with the drawings in the embodiment of the present application. Figure 1 , Figure 1 A flowchart of an audio processing method provided in an embodiment of the present application.

[0143] 101. Obtain a first audio to be tested and a target audio processing algorithm.

[0144] In the embodiment of the present application, the first audio to be tested may be any audio. The target audio processing algorithm may be any algorithm for processing audio, for example, the target audio processing algorithm is an algorithm for performing noise reduction processing on audio, for another example, the target audio processing algorithm is an audio codec algorithm, that is, the target audio processing algorithm can be used to encode and decode audio, for another example, the target audio processing algorithm includes an algorithm for performing noise reduction on audio and an algorithm for encoding audio.

[0145] In an implementation of obtaining the first audio to be tested, the audio processing device receives the first audio to be tested input by the user through an input component. The input component includes at least one of the following: a keyboard, a mouse, a touch screen, a touch pad, and an audio input device.

[0146] In another implementation of obtaining the first audio to be tested, the audio processing device receives the first audio to be tested sent by a terminal. The terminal may be any of the following: a mobile phone, a computer, a tablet computer, or a server.

[0147] In an implementation manner of acquiring a target audio processing algorithm, an audio processing device receives a target audio processing algorithm input by a user through an input component.

[0148] In another implementation of acquiring a target audio processing algorithm, the audio processing device receives the target audio processing algorithm sent by a terminal.

[0149] It should be understood that in the embodiment of the present application, the step of the audio processing device acquiring the first audio to be tested and the step of acquiring the target audio processing algorithm can be executed separately or simultaneously, and the present application does not limit this.

[0150] 102. Process the first audio to be tested by using a target audio processing algorithm to obtain first result data.

[0151] In an embodiment of the present application, the first result data is data generated by processing the first audio to be tested using the target audio processing algorithm. In one possible implementation, the first result data includes the first result audio. The first result audio is the audio obtained by processing the first audio to be tested using the target audio processing algorithm. In another possible implementation, the first result data includes the time consumed by processing the first audio to be tested using the target audio processing algorithm. In yet another possible implementation, the first result data includes the amount of computation consumed by processing the first audio to be tested using the target audio processing algorithm.

[0152] 103. According to the speech in the first result audio, determine the processing effect of the target audio processing algorithm on the speech to obtain a first evaluation result.

[0153] In an embodiment of the present application, the first evaluation result characterizes the processing effect of the target audio processing algorithm on the speech. The first result audio is the audio obtained by processing the first audio to be tested by the target audio processing algorithm. Therefore, the speech in the first result audio can be used to determine the processing effect of the target audio processing algorithm on the speech. In one possible implementation, the audio processing device determines the speech intelligibility of the speech in the first result audio, and determines the first evaluation result according to the speech intelligibility of the speech in the first result audio. At this time, the speech intelligibility of the speech in the first result audio can characterize the processing effect of the target audio processing algorithm on the speech. In another possible implementation, the audio processing device determines the speech intelligibility of the speech in the first result audio, and determines the first evaluation result according to the signal-to-noise ratio of the speech in the first result audio. At this time, the signal-to-noise ratio of the speech in the first result audio can characterize the processing effect of the target audio processing algorithm on the speech. In another possible implementation, the audio processing device determines the objective speech quality evaluation index (perceptual evaluation of speech quality, PESQ) of the first result audio. The first evaluation result is determined according to the PESQ of the first result audio. At this time, the PESQ in the first result audio can represent the processing effect of the target audio processing algorithm on the speech.

[0154] 104. Determine the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data to obtain a second evaluation result.

[0155] In an embodiment of the present application, the second evaluation result is based on the non-speech data in the first result data to determine the processing effect of the target audio processing algorithm on the audio. The non-speech data in the first result data includes data other than the speech in the first result audio. In one possible implementation, the non-speech data in the first result data is data other than the speech in the first result audio. At this time, the audio processing device can determine the processing effect of the target audio processing algorithm on the non-speech in the audio based on the non-speech data in the first result data, and then can determine the processing effect of the target audio processing algorithm on the audio based on the processing effect of the target audio processing algorithm on the non-speech in the audio, and obtain the second evaluation result. For example, the processing effect of the target audio processing algorithm on the non-speech in the audio is the noise reduction effect of the target audio processing algorithm on the non-speech in the audio. At this time, the second evaluation result characterizes the noise reduction effect of the target audio processing algorithm on the non-speech in the audio.

[0156] In another possible implementation, the non-speech data in the first result data is the time consumed by the target audio processing algorithm to process the first audio to be tested. At this time, the processing speed of the target audio processing algorithm on the audio can be determined based on the non-speech data in the first result data. The second evaluation result represents the processing speed of the target audio processing algorithm on the audio.

[0157] In another possible implementation, the non-speech data in the first result data is the amount of computation consumed by processing the first audio to be tested using the target audio processing algorithm. At this time, the amount of computation consumed by the target audio processing algorithm to process the audio can be determined based on the non-speech data in the first result data. The second evaluation result represents the amount of computation consumed by the target audio processing algorithm to process the audio.

[0158] 105. Determine the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result.

[0159] The audio processing device determines the processing effect of the target audio processing algorithm on the audio based on the first evaluation result and the second evaluation result, and can determine the processing effect of the target audio processing algorithm on the audio by utilizing the speech in the first result data and the non-speech data in the first result data.

[0160] In one possible implementation, the audio processing device uses the first evaluation result and the second evaluation result as the processing effect of the target audio processing algorithm on the audio.

[0161] In another possible implementation, the audio processing device performs a weighted summation on the first evaluation result and the second evaluation result to determine the processing effect of the target audio processing algorithm on the audio.

[0162] In an embodiment of the present application, after acquiring the first audio to be tested and the target audio processing algorithm, the audio processing device processes the first audio to be tested using the target audio processing algorithm to obtain first result data, wherein the first result data includes the first result audio obtained by processing the first audio to be tested using the target audio processing algorithm. The audio processing device then determines the processing effect of the target audio processing algorithm on the speech based on the speech in the first result audio to obtain a first evaluation result, and determines the processing effect of the target audio processing algorithm on the audio based on the non-speech data in the first result data to obtain a second evaluation result. Finally, based on the first evaluation result and the second evaluation result, the processing effect of the target audio processing algorithm on the audio is determined, and the processing effect of the target audio processing algorithm on the audio can be determined based on the speech in the first result data and the non-speech data in the first result data, thereby improving the accuracy of the evaluation of the processing effect of the target audio processing algorithm on the audio.

[0163] As an optional implementation, the audio processing device performs the following steps during the process of executing step 103:

[0164] 201. Obtain acoustic features of the first result audio by detecting the acoustic features of the first result audio.

[0165] In the embodiment of the present application, the acoustic feature refers to the physical quantity of the audio acoustic characteristic, such as the frequency, sound pressure level, and loudness of the audio. The audio processing device can obtain the acoustic feature of the first result audio by detecting the acoustic feature of the first result audio.

[0166] 202. According to the acoustic characteristics, the second evaluation result is obtained.

[0167] In one possible implementation, the acoustic feature includes a sound pressure level. The audio processing device determines a sound pressure level dynamic range according to a maximum value of the sound pressure level and a minimum value of the sound pressure level, wherein the sound pressure level dynamic range is a dynamic range of the sound pressure level of the first result audio. The second evaluation result is obtained according to the sound pressure level dynamic range.

[0168] In another possible implementation, the acoustic feature includes sound loudness. The audio processing device determines a loudness dynamic range according to a maximum loudness value and a minimum loudness value, wherein the loudness dynamic range is a dynamic range of loudness of the first result audio. The second evaluation result is obtained according to the loudness dynamic range.

[0169] In this embodiment, the audio processing device detects the acoustic features of the first result audio to obtain the acoustic features of the first result audio, and obtains the second evaluation result based on the acoustic features, so as to determine the processing effect of the target audio processing algorithm on the audio based on the non-speech data in the first result data.

[0170] As an optional implementation, the audio processing device performs the following steps during the execution of step 103: when the non-speech data includes a target duration, determining a second evaluation result based on time. The target duration represents the time consumed by the target audio processing algorithm to obtain the first result audio. In this implementation, the audio processing device can evaluate the performance of the audio processing algorithm based on the target duration to obtain a second evaluation result.

[0171] As an optional implementation, the audio processing device performs the following steps during the execution of step 103: when the non-speech data includes a calculation amount, determining the second evaluation result according to the calculation amount. The above-mentioned target calculation amount is the calculation amount generated by obtaining the first result audio through the target audio processing algorithm. In this implementation, the audio processing device can evaluate the performance of the audio processing algorithm according to the calculation amount to obtain the second evaluation result.

[0172] As an optional implementation, the audio processing device performs the following steps during the process of executing step 102:

[0173] 301. Determine the quality of the speech in the first result audio.

[0174] In a possible implementation manner, the audio processing device determines the quality of speech in the first result audio by determining the PESQ of the first result audio.

[0175] In another possible implementation manner, the audio processing device determines the quality of the speech in the first result audio by determining the signal-to-noise ratio of the first result audio.

[0176] In yet another possible implementation, the audio processing device determines the quality of speech in the first result audio by determining a mean opinion score (MOS) of the first result audio.

[0177] 302. Determine the processing effect of the target audio processing algorithm on the speech based on the quality, and obtain the first evaluation result.

[0178] In this implementation, after determining the quality of the speech in the first result audio, the audio processing device determines the processing effect of the target audio processing algorithm on the speech according to the quality to obtain a first evaluation result.

[0179] As an optional implementation, the audio processing device performs the following steps during the execution of step 102: determining the speech intelligibility of the speech in the first result audio. According to the speech intelligibility, determining the processing effect of the target audio processing algorithm on the speech, and obtaining the first evaluation result. In this implementation, after determining the speech intelligibility of the speech in the first result audio, the audio processing device determines the processing effect of the target audio processing algorithm on the speech according to the speech intelligibility, and obtains the first evaluation result.

[0180] As an optional implementation manner, before executing step 103, the audio processing device further executes the following steps:

[0181] 401. Obtain voice audio.

[0182] The above-mentioned voice audio is an audio including voice, and the above-mentioned first audio to be tested is an audio obtained based on the above-mentioned voice audio. In one possible implementation, the first audio to be tested is voice audio. In another possible implementation, music audio is obtained, wherein the sound in the music audio is music. By mixing the voice audio, the music audio and the noise, an audio that meets the preset requirements is obtained as the first audio to be tested. For example, the preset requirement is that the signal-to-noise ratio is a target signal-to-noise ratio, then the signal-to-noise ratio of the first audio to be tested is the target signal-to-noise ratio.

[0183] After acquiring the voice audio, the audio processing device performs the following steps in the process of executing step 103:

[0184] 402. When the energy of the speech audio is less than the energy of the first result audio, determining the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, the noise in the first audio to be tested remains in the first result audio.

[0185] Since the speech in the first audio to be tested is the same as the speech in the speech audio, the energy of the first result audio is greater than the energy of the speech audio, which means that in addition to the signal in the speech audio, the first result audio also contains noise, that is, there is noise in the first result audio obtained after being processed by the target audio processing algorithm, that is, the target audio processing algorithm does not completely remove the noise in the speech of the first audio to be tested. Therefore, the audio processing device can determine that the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, the noise in the first audio to be tested remains in the first result audio.

[0186] 403. When the energy of the speech audio is greater than the energy of the first result audio, determining the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, speech distortion exists in the first audio to be tested.

[0187] Since the speech in the first audio to be tested is the same as the speech in the speech audio, and the energy of the first result audio is less than the energy of the speech audio, it means that after being processed by the target audio processing algorithm, the speech in the first audio to be tested is reduced, that is, the processing by the target audio processing algorithm causes the speech in the first audio to be tested to be distorted. Therefore, the audio processing device can determine that the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, the speech in the first audio to be tested is distorted.

[0188] In this embodiment, the first audio to be tested is an audio obtained based on the speech audio, so the speech in the first audio to be tested is the same as the speech in the speech audio. Therefore, the audio processing device uses the speech audio as a reference to measure the processing effect of the target audio processing algorithm on the speech, and can obtain a first evaluation result.

[0189] As an optional implementation, the audio processing device performs the following steps during the process of executing step 105:

[0190] 501. Obtain a first reference range and a second reference range.

[0191] In the embodiment of the present application, the first reference range is the normal range of the first evaluation result, wherein when the first evaluation result is within the normal range, the first evaluation result is normal, and when the first evaluation result is outside the normal range, the first evaluation result is abnormal, which means that the target audio processing algorithm has an abnormality in processing the first audio to be tested. The second reference range is the normal range of the second evaluation result, wherein when the second evaluation result is within the normal range, the second evaluation result is normal, and when the second evaluation result is outside the normal range, the second evaluation result is abnormal, which means that the target audio processing algorithm has an abnormality in processing the second audio to be tested.

[0192] In one possible implementation, the audio processing device obtains at least one third evaluation result, wherein the third evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the speech according to the speech in the second result audio, the second result audio is the audio obtained by processing the second audio to be tested by the target audio processing algorithm, and the first audio to be tested is different from the second audio to be tested. The third evaluation result may be a result obtained by determining the first audio processing algorithm to be tested according to the speech in the third result audio, wherein the third result audio is the audio obtained by processing the audio by the first audio processing algorithm to be tested. The first evaluation result and at least one third evaluation result are arranged in order from large to small to obtain an arrangement result. The first reference range is determined according to the first preset percentile in the arrangement result and the second preset percentile in the arrangement result.

[0193] Optionally, a first reference value corresponding to a first preset quantile is determined from the arrangement result. A second reference value corresponding to a second preset quantile is determined from the arrangement result. A first reference range is determined based on the first reference value and the second reference value. Optionally, the first reference value is less than the second reference value, and the audio processing device uses the first reference value as the minimum value of the first reference range and the second reference value as the maximum value of the first reference range. Optionally, the second reference value is less than the first reference value, and the audio processing device uses the second reference value as the minimum value of the first reference range and the first reference value as the maximum value of the first reference range. Optionally, the first preset quantile is 1 / 4 and the second preset quantile is 3 / 4.

[0194] In this implementation, the first evaluation result and the third evaluation result are the same evaluation index. For example, the first evaluation result and the third evaluation result both represent speech intelligibility. For another example, the first evaluation result and the third evaluation result both represent the degree of speech distortion. However, the first evaluation result and the third evaluation result are obtained through different evaluation processes, wherein the different evaluation processes include processing different audios with the same audio processing algorithm, processing the same audio with different audio processing algorithms, and processing different audios with different audio processing algorithms. By statistically analyzing the first evaluation result and at least one third evaluation result, a first reference range of the first evaluation result can be determined.

[0195] 502. When the first evaluation result is outside the first reference range, or the second evaluation result is outside the second reference range, determining the processing effect of the target audio processing algorithm on the audio includes that the target audio processing algorithm processes the target class audio abnormally, and the target class audio is audio that meets the preset requirements.

[0196] In this implementation, the audio processing device determines whether the first evaluation result is normal based on the first reference range, and further determines whether the target audio processing algorithm processes the target class audio normally, thereby determining the boundary of the target audio processing algorithm.

[0197] As an optional implementation, the audio processing device may determine the second reference range in the same manner as the first reference range.

[0198] As an optional implementation, the audio processing device performs the following steps during the process of executing step 105:

[0199] 601. Obtain the fourth evaluation result and the fifth evaluation result.

[0200] In the embodiment of the present application, the fourth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the speech according to the speech in the fourth result audio, and the fourth result audio is obtained by processing the first audio to be tested by using the second audio processing algorithm to be evaluated, and the second audio processing algorithm to be evaluated is different from the target audio processing algorithm. In other words, the first evaluation result and the fourth evaluation result are obtained by processing the first audio to be tested by using different audio processing algorithms.

[0201] The fifth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the audio based on the non-speech data in the second result data, and the second result data is obtained by processing the first audio to be tested using the second audio processing algorithm to be evaluated. In other words, the second evaluation result and the fifth evaluation result are obtained by processing the first audio to be tested using different audio processing algorithms.

[0202] 602. Aggregate the first evaluation result and the fourth evaluation result to obtain a first aggregated result.

[0203] By aggregating the first evaluation result and the fourth evaluation result, it is possible to aggregate the evaluation results obtained by processing the first audio to be tested using different audio processing algorithms. For example, the first evaluation result and the fourth evaluation result both represent speech intelligibility. By aggregating the first evaluation result and the fourth evaluation result to obtain a first aggregated result, the difference in speech intelligibility after being processed by different audio processing algorithms can be determined through the first aggregated result. For another example, the first evaluation result and the fourth evaluation result both represent the degree of speech distortion. By aggregating the first evaluation result and the fourth evaluation result to obtain a first aggregated result, the difference in the degree of speech distortion after being processed by different audio processing algorithms can be determined through the first aggregated result.

[0204] 603. Aggregate the second evaluation result and the fifth evaluation result to obtain a second aggregation result.

[0205] By aggregating the second evaluation result and the fifth evaluation result, it is possible to aggregate the evaluation results obtained by processing the first audio to be tested using different audio processing algorithms. For example, the second evaluation result and the fifth evaluation result both characterize the dynamic range of the sound pressure level. By aggregating the second evaluation result and the fifth evaluation result to obtain a second aggregated result, the difference in the dynamic range of the sound pressure level after being processed by different audio processing algorithms can be determined through the second aggregated result. For another example, the second evaluation result and the fifth evaluation result both characterize the dynamic range of loudness. By aggregating the second evaluation result and the fifth evaluation result to obtain a second aggregated result, the difference in the dynamic range of loudness after being processed by different audio processing algorithms can be determined through the second aggregated result.

[0206] 604. Determine, according to the first aggregation result and the second aggregation result, a difference between a processing effect of the target audio processing algorithm and a processing effect of the second audio processing algorithm to be evaluated.

[0207] Specifically, the difference between the processing effect of the target audio processing algorithm on speech and the processing effect of the second audio processing algorithm to be tested can be characterized according to the first aggregation result. The difference between the processing effect of the target audio processing algorithm determined based on non-speech data and the processing effect of the second audio processing algorithm to be tested can be characterized according to the second aggregation result.

[0208] In this embodiment, the audio processing device can aggregate the evaluation results obtained by processing the first audio to be tested using different audio processing algorithms by aggregating the first evaluation result and the fourth evaluation result to obtain a first aggregated result, thereby characterizing the difference between the processing effect of the target audio processing algorithm on speech and the processing effect of the second audio processing algorithm to be tested on speech. The audio processing device can aggregate the evaluation results obtained by processing the first audio to be tested using different audio processing algorithms by aggregating the second evaluation result and the fifth evaluation result to obtain a second aggregated result, thereby characterizing the difference between the processing effect of the target audio processing algorithm determined based on non-speech data and the processing effect of the second audio processing algorithm to be tested. Therefore, the difference between the processing effect of the target audio processing algorithm and the processing effect of the second audio processing algorithm to be tested can be determined based on the first aggregated result and the second aggregated result.

[0209] As an optional implementation, the audio processing device performs the following steps during the process of executing step 105:

[0210] 701. Obtain the sixth evaluation result and the seventh evaluation result.

[0211] In the embodiment of the present application, the sixth evaluation result is the result obtained by determining the processing effect of the target audio processing algorithm on the speech in the fifth result audio, and the fifth result audio is obtained by processing the second audio to be tested using the target audio processing algorithm, and the second audio to be tested is different from the first audio to be tested. In other words, the first evaluation result and the seventh evaluation result are obtained by processing different audios using the same audio processing algorithm.

[0212] The seventh evaluation result is obtained by determining the processing effect of the target audio processing algorithm on the audio based on the non-speech data in the third result data, and the third result data is obtained by processing the second audio to be tested using the target audio processing algorithm. In other words, the second evaluation result and the seventh evaluation result are obtained by processing different audios using the same audio processing algorithm.

[0213] 702. Aggregate the first evaluation result and the sixth evaluation result to obtain a third aggregation result.

[0214] By aggregating the first evaluation result and the sixth evaluation result, it is possible to aggregate the evaluation results obtained by processing different audios using the same audio processing algorithm. For example, the first evaluation result and the sixth evaluation result both represent speech intelligibility. By aggregating the first evaluation result and the sixth evaluation result, a third aggregate result is obtained. The third aggregate result can be used to determine the difference in speech intelligibility after different audios are processed by the same audio processing algorithm. For another example, the first evaluation result and the sixth evaluation result both represent the degree of speech distortion. By aggregating the first evaluation result and the sixth evaluation result, a third aggregate result is obtained. The third aggregate result can be used to determine the difference in the degree of speech distortion after different audios are processed by the same audio processing algorithm.

[0215] 703. Aggregate the second evaluation result and the seventh evaluation result to obtain a fourth aggregation result.

[0216] By aggregating the second evaluation result and the seventh evaluation result, it is possible to aggregate the evaluation results obtained by processing different audios using the same audio processing algorithm. For example, the second evaluation result and the seventh evaluation result both characterize the dynamic range of the sound pressure level. By aggregating the second evaluation result and the seventh evaluation result, a fourth aggregate result is obtained. The fourth aggregate result can be used to determine the difference in the dynamic range of the sound pressure level after different audios are processed by the same audio processing algorithm. For another example, the second evaluation result and the seventh evaluation result both characterize the dynamic range of loudness. By aggregating the second evaluation result and the seventh evaluation result, a fourth aggregate result is obtained. The fourth aggregate result can be used to determine the difference in the dynamic range of loudness after different audios are processed by the same audio processing algorithm.

[0217] 704. Determine the processing effect of the target audio processing algorithm on different audios according to the third aggregation result and the fourth aggregation result.

[0218] Specifically, the difference between the processing effect of the target audio processing algorithm on speech and the processing effect of the second audio processing algorithm to be tested can be characterized according to the first aggregation result. The difference between the processing effect of the target audio processing algorithm determined based on non-speech data and the processing effect of the second audio processing algorithm to be tested can be characterized according to the second aggregation result.

[0219] In this embodiment, the audio processing device can aggregate the evaluation results obtained by processing different audios using the same audio processing algorithm by aggregating the first evaluation result and the sixth evaluation result, and obtain a third aggregate result, thereby characterizing the processing effect of the target audio processing algorithm on the speech in different audios. The audio processing device can aggregate the evaluation results obtained by processing the first audio to be tested using different audio processing algorithms by aggregating the second evaluation result and the fifth evaluation result, and obtain a fourth aggregate result, thereby characterizing the processing effect of the target audio processing algorithm determined based on non-speech data on the speech in different audios. Therefore, the processing effect of the target audio processing algorithm on different audios can be determined based on the third aggregate result and the fourth aggregate result.

[0220] As an optional implementation, the audio processing device performs the following steps during the process of executing step 105:

[0221] 801. When the number of the first evaluation results is greater than 1, weighted fusion is performed on the first evaluation results to obtain an eighth evaluation result.

[0222] For example, the first evaluation result includes evaluation result a and evaluation result b, wherein evaluation result a represents speech intelligibility, and evaluation result b represents the degree of speech distortion, and the eighth evaluation result is obtained by weighted fusion of evaluation result a and evaluation result b.

[0223] 802. Determine the processing effect of the target audio processing algorithm on the audio according to the eighth evaluation result and the second evaluation result.

[0224] In this manner, when the number of the first evaluation results is greater than 1, the audio processing device first performs weighted fusion on the first evaluation results to obtain an eighth evaluation result, thereby reducing the number of evaluation results. Then, based on the eighth evaluation result and the second evaluation result, the processing effect of the target audio processing algorithm on the audio is determined, thereby reducing the amount of data processing.

[0225] As an optional implementation, the audio processing device performs the following steps during the process of executing step 105:

[0226] 901. When the number of the second evaluation results is greater than 1, weighted fusion is performed on the second evaluation results to obtain a ninth evaluation result.

[0227] For example, the second evaluation result includes evaluation result c and evaluation result d, wherein evaluation result c represents the dynamic range of the sound pressure level, and evaluation result d represents the dynamic range of the loudness. The ninth evaluation result is obtained by weighted fusion of evaluation result c and evaluation result d.

[0228] 902. Determine the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the ninth evaluation result.

[0229] In this manner, when the number of the second evaluation results is greater than 1, the audio processing device first performs weighted fusion on the second evaluation results to obtain a ninth evaluation result, thereby reducing the number of evaluation results. Then, based on the ninth evaluation result and the first evaluation result, the processing effect of the target audio processing algorithm on the audio is determined, thereby reducing the amount of data processing.

[0230] As an optional implementation, the audio processing device obtains the target audio processing algorithm by performing the following steps:

[0231] 1001. Obtain target configuration parameters of the target audio processing algorithm.

[0232] In an embodiment of the present application, the target configuration parameters include at least one candidate audio processing algorithm required to obtain the target audio processing algorithm, a coupling relationship between at least one candidate audio processing algorithm, and at least one candidate configuration parameter of at least one candidate audio processing algorithm. For example, at least one candidate audio processing algorithm includes audio processing algorithm a and audio processing algorithm b, the coupling relationship includes first processing with audio processing algorithm a and then processing with audio processing algorithm b, and at least one candidate configuration parameter includes configuration parameters of audio processing algorithm a and configuration parameters of audio processing algorithm b.

[0233] In an implementation method of obtaining target configuration parameters of a target audio processing algorithm, an audio processing device receives the target configuration parameters input by a user through an input component.

[0234] In another implementation manner of acquiring target configuration parameters of a target audio processing algorithm, the audio processing device receives the target configuration parameters sent by a terminal.

[0235] 1002. Configure the at least one candidate audio processing algorithm according to the at least one candidate configuration parameter and the coupling relationship to obtain the target audio processing algorithm.

[0236] In this implementation, after obtaining the target configuration parameters of the target audio processing algorithm, the audio processing device configures at least one candidate audio processing algorithm according to at least one candidate configuration parameter and the coupling relationship to obtain the target audio processing algorithm. In this way, the target audio processing algorithm that meets the requirements can be configured according to the requirements. Specifically, the target configuration parameters are configured according to the actual requirements, and then the target audio processing algorithm that meets the requirements can be obtained.

[0237] See also Figure 2 , Figure 2A flowchart of another audio processing method provided in an embodiment of the present application is shown below. Figure 2 As shown, the whole process includes a data preparation stage, a data processing stage, an evaluation and processing stage, and a result analysis stage. The data preparation stage includes generating a test data set based on a basic data set, and preparing an audio processing algorithm, wherein the basic data set includes a speech set, a noise set, and a music set. The audio in the speech set is speech audio, the data in the noise set is noise, and the audio in the music set is music audio. By mixing speech audio, music audio, and noise, an audio that meets the preset requirements can be obtained (the specific implementation process can be seen in step 401), and then a test data set can be obtained, that is, the audio in the test data set is the audio obtained by mixing speech audio, music audio, and noise. It should be understood that the test data set includes at least two different audios. The audio processing algorithm is any algorithm for processing audio, and the number of audio processing algorithms is at least one.

[0238] In the data processing stage, configuration and task scheduling are performed based on the test data set and the audio processing algorithm to obtain an operation result data set, wherein the configuration and task scheduling include configuring the audio processing algorithm based on the configuration parameters of the audio processing algorithm to obtain the audio processing algorithm required for the test (the specific implementation process can be seen in steps 1001 and 1002), and the operation result data set includes the result data obtained by processing the audio in the test data set using the audio processing algorithm required for the test (specifically, see the implementation process of obtaining the first result data in step 102). In the evaluation processing stage, based on the operation result data set and the test data set, evaluation algorithm selection and calculation and outlier detection calculation are performed. The implementation process of evaluation algorithm selection and calculation can be seen in step 103 to obtain the first evaluation result and step 104 to obtain the second evaluation result. The implementation process of outlier detection calculation can be seen in steps 501 and 502. In the result analysis stage, evaluation result analysis, evaluation report and algorithm optimization are performed. In the implementation process of evaluation result analysis, reference can be made to steps 601 to 604, steps 701 to 704, steps 801 to 802, and steps 901 to 902. Evaluation report and algorithm optimization refer to that an evaluation report can be obtained through evaluation result analysis, and the audio processing algorithm can also be configured based on the evaluation result analysis.

[0239] pass Figure 2 The process of the audio processing method shown in the figure can evaluate the processing effect of the audio processing algorithm and provide guidance for the optimization of the audio processing algorithm. In a possible scenario, the audio in the test data set is audio that matches the audio in the on-demand platform. Figure 2The audio processing method can determine which audio processing algorithm is suitable for processing the audio of the on-demand platform, and how to optimize the audio processing algorithm to improve the processing effect of the audio of the on-demand platform. In another possible scenario, the audio in the test data set is audio that matches the audio in the live broadcast platform. Figure 2 The audio processing method can determine which audio processing algorithm is suitable for processing the audio of the live broadcast platform, and how to optimize the audio processing algorithm can improve the processing effect of the audio of the live broadcast platform.

[0240] See also Figure 3 , Figure 3 A flowchart of another audio processing method provided in an embodiment of the present application is shown below. Figure 3 As shown, by performing data enhancement configuration on the speech set, the noise set, and the music set, a test data set can be obtained. Specifically, based on the signal-to-noise ratio, sound pressure level, reverberation, cropping, resampling, etc., the speech audio, music audio, and noise are mixed to obtain audio that meets the preset requirements (the implementation process of this step can be seen in step 401), which is used as the audio in the test data set. Figure 3 The music detection, dynamic gain control, echo cancellation, and noise suppression in the algorithm to be tested are all uses of the algorithm to be tested, where the algorithm to be tested is an audio processing algorithm. By calling the API of the algorithm to be tested, the corresponding algorithm to be tested can be called. For example, by calling the API of the audio processing algorithm for music detection, the audio processing algorithm for music detection can be called.

[0241] After calling the API through the algorithm to be tested and calling the corresponding algorithm to be tested, the algorithm to be tested configuration and task scheduling can be performed based on the called algorithm to be tested, and the algorithm to be tested can be obtained (the specific implementation process can be seen in steps 1001 and 1002). Then, through the call and calculation of the algorithm to be tested, the test data set is processed by the algorithm to be tested to obtain the operation result data set (specifically, see the implementation process of obtaining the first result data in step 102). The test data set and the operation result data set are then stored and archived for subsequent query.

[0242] After obtaining the running result data set, the algorithm used to evaluate the processing effect of the algorithm to be tested is determined through the configuration and selection of the evaluation algorithm. For example, the processing effect of the algorithm to be tested is determined through the algorithm with parameter evaluation and the algorithm without parameter evaluation. After determining the algorithm used to evaluate the processing effect of the algorithm to be tested, the algorithm used to evaluate the processing effect of the algorithm to be tested is used to process the audio data. Specifically, the algorithm used to evaluate the processing effect of the algorithm to be tested is used to process the audio in the running result data set, and then the evaluation algorithm starts to calculate, and the evaluation result is output through the evaluation result output (for details, please refer to the implementation process of obtaining the first evaluation result in step 103 and the implementation process of obtaining the second evaluation result in step 104).

[0243] Through the outlier and abnormal audio selection method, the evaluation results are calculated and detected for outliers (see step 501 and step 502 for details). According to the evaluation results, the results of outlier calculation and detection, the evaluation results are compared and analyzed (see steps 601 to 604, steps 701 to 704, steps 801 to 802, and steps 901 to 902 in the implementation process). Based on the results of comparison and analysis, algorithm selection and optimization can be performed. Specifically, the algorithm to be tested with good processing effect is selected, and the configuration parameters of the algorithm to be tested are optimized. At the same time, the results of comparison and analysis can also be stored and archived, wherein the results of comparison and analysis include PESQ, virtual speech quality target listener (ViSQOL), short-time objective intelligibility (STOI), noise residual, sound pressure level, dynamic range of loudness, loudness...

[0244] Those skilled in the art will appreciate that, in the above method of specific implementation, the order in which the steps are written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of the steps should be determined by their functions and possible internal logic.

[0245] The method of the embodiment of the present application is described in detail above, and the device of the embodiment of the present application is provided below.

[0246] See also Figure 4 , Figure 4 This is a structural diagram of an audio processing device provided in an embodiment of the present application. The audio processing device 1 includes: an acquisition unit 11, a processing unit 12, and a determination unit 13. Specifically:

[0247] An acquisition unit 11 is used to acquire a first audio to be tested and a target audio processing algorithm;

[0248] The processing unit 12 is used to process the first audio to be tested by using a target audio processing algorithm to obtain first result data, where the first result data includes a first result audio, where the first result audio is an audio obtained by processing the first audio to be tested by using the target audio processing algorithm;

[0249] A determination unit 13, configured to determine, based on the speech in the first result audio, a processing effect of the target audio processing algorithm on the speech, and obtain a first evaluation result;

[0250] The determining unit 13 is used to determine the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data, so as to obtain a second evaluation result;

[0251] The determining unit 13 is used to determine the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result.

[0252] In combination with any implementation manner of the present application, the determining unit 13 is specifically configured to:

[0253] By detecting the acoustic features of the first result audio, the acoustic features of the first result audio are obtained;

[0254] The second evaluation result is obtained according to the acoustic feature.

[0255] In combination with any embodiment of the present application, the acoustic feature includes one or more of sound pressure level and loudness;

[0256] The determining unit 13 is specifically configured to:

[0257] In the case where the acoustic feature includes a sound pressure level, determining a sound pressure level dynamic range according to a maximum value of the sound pressure level and a minimum value of the sound pressure level, the sound pressure level dynamic range being a dynamic range of the sound pressure level of the first result audio;

[0258] Obtaining the second evaluation result according to the sound pressure level dynamic range;

[0259] In the case where the acoustic feature includes loudness, determining a loudness dynamic range according to a maximum value of the loudness and a minimum value of the loudness, wherein the loudness dynamic range is a dynamic range of the loudness of the first result audio;

[0260] The second evaluation result is obtained according to the loudness dynamic range.

[0261] In combination with any embodiment of the present application, the non-speech data includes one or more of a target duration and a target calculation amount, the target duration represents the time consumed by the target audio processing algorithm to obtain the first result audio, and the target calculation amount is the calculation amount generated by obtaining the first result audio through the target audio processing algorithm;

[0262] The determining unit 13 is specifically configured to:

[0263] In a case where the non-speech data includes the time, determining the second evaluation result according to the time;

[0264] In a case where the non-speech data includes the calculation amount, the second evaluation result is determined according to the calculation amount.

[0265] In combination with any implementation manner of the present application, the determining unit 13 is specifically configured to:

[0266] Determining the quality of the speech in the first result audio;

[0267] Determine, based on the quality, a processing effect of the target audio processing algorithm on the speech, and obtain the first evaluation result;

[0268] Alternatively, determining speech intelligibility of speech in the first result audio;

[0269] According to the speech intelligibility, the processing effect of the target audio processing algorithm on the speech is determined to obtain the first evaluation result.

[0270] In combination with any embodiment of the present application, the acquisition unit 11 is further used to acquire voice audio, where the voice audio is audio including voice, and the first audio to be tested is audio obtained based on the voice audio;

[0271] The determining unit 13 is specifically configured to:

[0272] In a case where the energy of the speech audio is less than the energy of the first result audio, determining the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, the noise in the first audio to be tested remains in the first result audio;

[0273] When the energy of the speech audio is greater than the energy of the first result audio, determining the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, speech distortion exists in the first audio to be tested.

[0274] In combination with any embodiment of the present application, the acquisition unit 11 is specifically used for:

[0275] Acquire music audio, where the sound in the music audio is music;

[0276] By mixing the voice audio, the music audio and the noise, an audio satisfying a preset requirement is obtained as the first audio to be tested.

[0277] In combination with any implementation manner of the present application, the determining unit 13 is specifically configured to:

[0278] Obtaining a first reference range and a second reference range, wherein the first reference range is a normal range of the first evaluation result, and the second reference range is a normal range of the second evaluation result;

[0279] When the first evaluation result is outside the first reference range, or the second evaluation result is outside the second reference range, determining the processing effect of the target audio processing algorithm on the audio includes that the target audio processing algorithm processes the target class audio abnormally, and the target class audio is audio that meets the preset requirements.

[0280] In combination with any embodiment of the present application, the acquisition unit 11 is specifically used for:

[0281] Obtain at least one third evaluation result, wherein the third evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the speech according to the speech in the second result audio, the second result audio is the audio obtained by processing the second audio to be tested by the target audio processing algorithm, and the first audio to be tested is different from the second audio to be tested; the third evaluation result may be a result obtained by determining the first audio processing algorithm to be tested according to the speech in the third result audio, and the third result audio is the audio obtained by processing the audio by the first audio processing algorithm to be tested;

[0282] Arrange the first evaluation result and the at least one third evaluation result in order from largest to smallest to obtain an arrangement result;

[0283] The first reference range is determined according to a first preset quantile in the arrangement result and a second preset quantile in the arrangement result.

[0284] In combination with any implementation manner of the present application, the determining unit 13 is specifically configured to:

[0285] Determining a first reference value corresponding to the first preset quantile from the arrangement result;

[0286] Determining a second reference value corresponding to the second preset quantile from the arrangement result;

[0287] The first reference range is determined according to the first reference value and the second reference value.

[0288] In combination with any implementation manner of the present application, the determining unit 13 is specifically configured to:

[0289] Obtain a fourth evaluation result and a fifth evaluation result, wherein the fourth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the speech based on the speech in the fourth result audio, and the fourth result audio is obtained by processing the first audio to be evaluated by using the second audio processing algorithm to be evaluated, and the second audio processing algorithm to be evaluated is different from the target audio processing algorithm; the fifth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the audio based on the non-speech data in the second result data, and the second result data is obtained by processing the first audio to be tested by using the second audio processing algorithm to be evaluated;

[0290] Aggregating the first evaluation result and the fourth evaluation result to obtain a first aggregated result;

[0291] Aggregating the second evaluation result and the fifth evaluation result to obtain a second aggregated result;

[0292] According to the first aggregation result and the second aggregation result, a difference between a processing effect of the target audio processing algorithm and a processing effect of the second audio processing algorithm to be evaluated is determined.

[0293] In combination with any implementation manner of the present application, the determining unit 13 is specifically configured to:

[0294] Obtain a sixth evaluation result and a seventh evaluation result, wherein the sixth evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the speech based on the speech in the fifth result audio, and the fifth result audio is obtained by processing the second audio to be tested using the target audio processing algorithm, and the second audio to be tested is different from the first audio to be tested; the seventh evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the audio based on the non-speech data in the third result data, and the third result data is obtained by processing the second audio to be tested using the target audio processing algorithm;

[0295] Aggregating the first evaluation result and the sixth evaluation result to obtain a third aggregated result;

[0296] Aggregating the second evaluation result and the seventh evaluation result to obtain a fourth aggregated result;

[0297] The processing effect of the target audio processing algorithm on different audios is determined according to the third aggregation result and the fourth aggregation result.

[0298] In combination with any implementation manner of the present application, the determining unit 13 is specifically configured to:

[0299] When the number of the first evaluation results is greater than 1, weighted fusion is performed on the first evaluation results to obtain an eighth evaluation result;

[0300] Determining, according to the eighth evaluation result and the second evaluation result, a processing effect of the target audio processing algorithm on the audio;

[0301] Alternatively, when the number of the second evaluation results is greater than 1, weighted fusion is performed on the second evaluation results to obtain a ninth evaluation result;

[0302] The processing effect of the target audio processing algorithm on the audio is determined according to the first evaluation result and the eighth evaluation result.

[0303] In combination with any embodiment of the present application, the acquisition unit 11 is specifically used for:

[0304] Acquire target configuration parameters of the target audio processing algorithm, where the target configuration parameters include at least one candidate audio processing algorithm required to obtain the target audio processing algorithm, a coupling relationship between the at least one candidate audio processing algorithm, and at least one candidate configuration parameter of the at least one candidate audio processing algorithm;

[0305] The at least one candidate audio processing algorithm is configured according to the at least one candidate configuration parameter and the coupling relationship to obtain the target audio processing algorithm.

[0306] In an embodiment of the present application, after acquiring the first audio to be tested and the target audio processing algorithm, the audio processing device processes the first audio to be tested using the target audio processing algorithm to obtain first result data, wherein the first result data includes the first result audio obtained by processing the first audio to be tested using the target audio processing algorithm. The audio processing device then determines the processing effect of the target audio processing algorithm on the speech based on the speech in the first result audio to obtain a first evaluation result, and determines the processing effect of the target audio processing algorithm on the audio based on the non-speech data in the first result data to obtain a second evaluation result. Finally, based on the first evaluation result and the second evaluation result, the processing effect of the target audio processing algorithm on the audio is determined, and the processing effect of the target audio processing algorithm on the audio can be determined based on the speech in the first result data and the non-speech data in the first result data, thereby improving the accuracy of the evaluation of the processing effect of the target audio processing algorithm on the audio.

[0307] In some embodiments, the functions or modules included in the device provided in the embodiments of the present application can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments. For the sake of brevity, it will not be repeated here.

[0308] Figure 5 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application. The electronic device 2 includes a processor 21 and a memory 22. Optionally, the electronic device 2 also includes an input device 23 and an output device 24. The processor 21, the memory 22, the input device 23 and the output device 24 are coupled via a connector, and the connector includes various interfaces, transmission lines or buses, etc., which are not limited in the embodiments of the present application. It should be understood that in each embodiment of the present application, coupling refers to mutual connection in a specific manner, including direct connection or indirect connection through other devices, for example, it can be connected through various interfaces, transmission lines, buses, etc.

[0309] The processor 21 may include one or more processors, for example, one or more central processing units (CPUs). When the processor is a CPU, the CPU may be a single-core CPU or a multi-core CPU. Optionally, the processor 21 may be a processor group consisting of multiple CPUs, and the multiple processors are coupled to each other via one or more buses. Optionally, the processor may also be other types of processors, etc., which are not limited in the embodiments of the present application.

[0310] The memory 22 can be used to store computer program instructions and various computer program codes including the program code for executing the program code of the present application. Optionally, the memory includes but is not limited to random access memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM), or portable read only memory (CD-ROM), which is used for related instructions and data.

[0311] The input device 23 is used to input data and / or signals, and the output device 24 is used to output data and / or signals. The input device 23 and the output device 24 can be independent devices or an integrated device.

[0312] It can be understood that in the embodiment of the present application, the memory 22 can be used not only to store relevant instructions, but also to store relevant data. The embodiment of the present application does not limit the specific data stored in the memory.

[0313] Understandably, Figure 5 Only a simplified design of an electronic device is shown. In practical applications, the electronic device may also include other necessary components, including but not limited to any number of input / output devices, processors, memories, etc., and all electronic devices that can implement the embodiments of the present application are within the protection scope of the present application.

[0314] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0315] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. Those skilled in the art can also clearly understand that the descriptions of the various embodiments of the present application have different focuses. For the convenience and brevity of description, the same or similar parts may not be repeated in different embodiments. Therefore, for parts not described or not described in detail in a certain embodiment, refer to the records of other embodiments.

[0316] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0317] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0318] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0319] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital versatile disc (DVD)), or a semiconductor medium (eg, a solid state disk (SSD)).

[0320] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by a computer program to instruct the relevant hardware, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned method embodiments. The aforementioned storage medium includes: a read-only memory (ROM) or a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

Claims

1. An audio processing method, characterized in that: The method comprises: Obtaining a first audio to be tested and a target audio processing algorithm; Processing the first audio to be tested by using a target audio processing algorithm to obtain first result data, where the first result data includes a first result audio, where the first result audio is audio obtained by processing the first audio to be tested by using the target audio processing algorithm; Determine, according to the speech in the first result audio, the processing effect of the target audio processing algorithm on the speech, and obtain a first evaluation result; Determining the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data to obtain a second evaluation result, wherein the non-speech data includes one or more of a target duration and a target calculation amount, wherein the target duration represents the time consumed by the target audio processing algorithm to obtain the first result audio, and the target calculation amount is the calculation amount generated by obtaining the first result audio through the target audio processing algorithm; determining the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data to obtain the second evaluation result includes: when the non-speech data includes the target duration, determining the second evaluation result according to the time; when the non-speech data includes the calculation amount, determining the second evaluation result according to the calculation amount; Determining the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result; Determining the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result includes: obtaining at least one third evaluation result, the third evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the speech according to the speech in the second result audio, the second result audio is the audio obtained by processing the second audio to be tested by the target audio processing algorithm, and the first audio to be tested is different from the second audio to be tested; the third evaluation result may be a result obtained by determining the first audio processing algorithm to be tested according to the speech in the third result audio, the third result audio is the audio obtained by processing the audio by the first audio processing algorithm to be tested; arranging the first evaluation result and the at least one third evaluation result in order from large to small to obtain an arrangement result; determining a first reference range according to a first preset quantile in the arrangement result and a second preset quantile in the arrangement result; when the first evaluation result is outside the first reference range, determining the processing effect of the target audio processing algorithm on the audio includes that the target audio processing algorithm has an abnormality in processing the target audio, and the target audio is audio that meets the preset requirements; The method of determining the processing effect of the target audio processing algorithm on the audio based on the first evaluation result and the second evaluation result may include: obtaining a fourth evaluation result and a fifth evaluation result, wherein the fourth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the speech based on the speech in the fourth result audio, and the fourth result audio is obtained by processing the first audio to be evaluated by the second audio processing algorithm to be evaluated, and the second audio processing algorithm to be evaluated is different from the target audio processing algorithm; the fifth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the audio based on the non-speech data in the second result data, and the second result data is obtained by processing the first audio to be evaluated by the second audio processing algorithm to be evaluated; aggregating the first evaluation result and the fourth evaluation result to obtain a first aggregated result; aggregating the second evaluation result and the fifth evaluation result to obtain a second aggregated result; and determining the difference between the processing effect of the target audio processing algorithm and the processing effect of the second audio processing algorithm to be evaluated based on the first aggregated result and the second aggregated result.

2. The method according to claim 1, characterized in that: The step of determining the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data to obtain a second evaluation result includes: By detecting the acoustic features of the first result audio, the acoustic features of the first result audio are obtained; The second evaluation result is obtained according to the acoustic feature.

3. The method according to claim 2, characterized in that The acoustic characteristics include one or more of sound pressure level and loudness; The obtaining the second evaluation result according to the acoustic feature includes: In the case where the acoustic feature includes a sound pressure level, determining a sound pressure level dynamic range according to a maximum value of the sound pressure level and a minimum value of the sound pressure level, the sound pressure level dynamic range being a dynamic range of the sound pressure level of the first result audio; Obtaining the second evaluation result according to the sound pressure level dynamic range; In the case where the acoustic feature includes loudness, determining a loudness dynamic range according to a maximum value of the loudness and a minimum value of the loudness, wherein the loudness dynamic range is a dynamic range of the loudness of the first result audio; The second evaluation result is obtained according to the loudness dynamic range.

4. The method according to claim 1, characterized in that: The step of determining, based on the speech in the first result audio, a processing effect of the target audio processing algorithm on the speech to obtain a first evaluation result includes: Determining the quality of the speech in the first result audio; Determine, based on the quality, a processing effect of the target audio processing algorithm on the speech, and obtain the first evaluation result; Alternatively, determining speech intelligibility of speech in the first result audio; According to the speech intelligibility, the processing effect of the target audio processing algorithm on the speech is determined to obtain the first evaluation result.

5. The method according to claim 1, characterized in that Before determining the processing effect of the target audio processing algorithm on the speech according to the speech in the first result audio and obtaining the first evaluation result, the method further includes: Acquire voice audio, where the voice audio is audio including voice, and the first audio to be tested is audio obtained based on the voice audio; The step of determining, based on the speech in the first result audio, a processing effect of the target audio processing algorithm on the speech to obtain a first evaluation result includes: In a case where the energy of the speech audio is less than the energy of the first result audio, determining the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, the noise in the first audio to be tested remains in the first result audio; When the energy of the speech audio is greater than the energy of the first result audio, determining the first evaluation result includes: after the first audio to be tested is processed by the target audio processing algorithm, speech distortion exists in the first audio to be tested.

6. The method according to claim 5, characterized in that The step of obtaining the first audio to be tested includes: Acquire music audio, where the sound in the music audio is music; By mixing the voice audio, the music audio and the noise, an audio satisfying a preset requirement is obtained as the first audio to be tested.

7. The method according to claim 6, characterized in that Determining the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result includes: obtaining a second reference range; When the second evaluation result is outside the second reference range, determining the processing effect of the target audio processing algorithm on the audio includes that the target audio processing algorithm processes the target class audio abnormally, and the target class audio is audio that meets the preset requirements.

8. The method according to claim 1, characterized in that: The determining the first reference range according to the first preset quantile in the arrangement result and the second preset quantile in the arrangement result includes: Determining a first reference value corresponding to the first preset quantile from the arrangement result; Determining a second reference value corresponding to the second preset quantile from the arrangement result; The first reference range is determined according to the first reference value and the second reference value.

9. The method according to claim 1, characterized in that: Determining the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result includes: Obtain a sixth evaluation result and a seventh evaluation result, wherein the sixth evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the speech based on the speech in the fifth result audio, and the fifth result audio is obtained by processing the second audio to be tested using the target audio processing algorithm, and the second audio to be tested is different from the first audio to be tested; the seventh evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the audio based on the non-speech data in the third result data, and the third result data is obtained by processing the second audio to be tested using the target audio processing algorithm; Aggregating the first evaluation result and the sixth evaluation result to obtain a third aggregated result; Aggregating the second evaluation result and the seventh evaluation result to obtain a fourth aggregated result; The processing effect of the target audio processing algorithm on different audios is determined according to the third aggregation result and the fourth aggregation result.

10. The method according to claim 1, characterized in that Determining the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result includes: When the number of the first evaluation results is greater than 1, weighted fusion is performed on the first evaluation results to obtain an eighth evaluation result; Determining, according to the eighth evaluation result and the second evaluation result, a processing effect of the target audio processing algorithm on the audio; Alternatively, when the number of the second evaluation results is greater than 1, weighted fusion is performed on the second evaluation results to obtain a ninth evaluation result; The processing effect of the target audio processing algorithm on the audio is determined according to the first evaluation result and the ninth evaluation result.

11. The method according to claim 1, characterized in that: Get the target audio processing algorithm, including: Acquire target configuration parameters of the target audio processing algorithm, where the target configuration parameters include at least one candidate audio processing algorithm required to obtain the target audio processing algorithm, a coupling relationship between the at least one candidate audio processing algorithm, and at least one candidate configuration parameter of the at least one candidate audio processing algorithm; The at least one candidate audio processing algorithm is configured according to the at least one candidate configuration parameter and the coupling relationship to obtain the target audio processing algorithm.

12. An audio processing device, characterized in that: The audio processing device comprises: An acquisition unit, used to acquire a first audio to be tested and a target audio processing algorithm; a processing unit, configured to process the first audio to be tested by using a target audio processing algorithm to obtain first result data, wherein the first result data includes a first result audio, and the first result audio is an audio obtained by processing the first audio to be tested by using the target audio processing algorithm; a determining unit, configured to determine, based on the speech in the first result audio, a processing effect of the target audio processing algorithm on the speech, and obtain a first evaluation result; The determining unit is used to determine the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data to obtain a second evaluation result, wherein the non-speech data includes one or more of a target duration and a target calculation amount, wherein the target duration represents the time consumed by the target audio processing algorithm to obtain the first result audio, and the target calculation amount is the calculation amount generated by obtaining the first result audio through the target audio processing algorithm; the determining the processing effect of the target audio processing algorithm on the audio according to the non-speech data in the first result data to obtain the second evaluation result includes: when the non-speech data includes the target duration, determining the second evaluation result according to the time; when the non-speech data includes the calculation amount, determining the second evaluation result according to the calculation amount; The determining unit is used to determine the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result; Determining the processing effect of the target audio processing algorithm on the audio according to the first evaluation result and the second evaluation result includes: obtaining at least one third evaluation result, the third evaluation result is a result obtained by determining the processing effect of the target audio processing algorithm on the speech according to the speech in the second result audio, the second result audio is the audio obtained by processing the second audio to be tested by the target audio processing algorithm, and the first audio to be tested is different from the second audio to be tested; the third evaluation result may be a result obtained by determining the first audio processing algorithm to be tested according to the speech in the third result audio, the third result audio is the audio obtained by processing the audio by the first audio processing algorithm to be tested; arranging the first evaluation result and the at least one third evaluation result in order from large to small to obtain an arrangement result; determining a first reference range according to a first preset quantile in the arrangement result and a second preset quantile in the arrangement result; when the first evaluation result is outside the first reference range, determining the processing effect of the target audio processing algorithm on the audio includes that the target audio processing algorithm has an abnormality in processing the target audio, and the target audio is audio that meets the preset requirements; The method of determining the processing effect of the target audio processing algorithm on the audio based on the first evaluation result and the second evaluation result may include: obtaining a fourth evaluation result and a fifth evaluation result, wherein the fourth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the speech based on the speech in the fourth result audio, and the fourth result audio is obtained by processing the first audio to be evaluated by the second audio processing algorithm to be evaluated, and the second audio processing algorithm to be evaluated is different from the target audio processing algorithm; the fifth evaluation result is a result obtained by determining the processing effect of the second audio processing algorithm to be evaluated on the audio based on the non-speech data in the second result data, and the second result data is obtained by processing the first audio to be evaluated by the second audio processing algorithm to be evaluated; aggregating the first evaluation result and the fourth evaluation result to obtain a first aggregated result; aggregating the second evaluation result and the fifth evaluation result to obtain a second aggregated result; and determining the difference between the processing effect of the target audio processing algorithm and the processing effect of the second audio processing algorithm to be evaluated based on the first aggregated result and the second aggregated result.

13. An electronic device, characterized in that: include: A processor and a memory, wherein the memory is used to store computer program codes, wherein the computer program codes include computer instructions, and when the processor executes the computer instructions, the electronic device executes the method according to any one of claims 1 to 11.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor, the processor is enabled to execute the method according to any one of claims 1 to 11.

15. A computer program product, characterized in that The computer program product comprises a computer program or instructions; when the computer program or instructions are run on a computer, the computer is caused to execute the method according to any one of claims 1 to 11.

Images