Audio recognition method, apparatus, medium, and computing device

Abstract

Embodiments of the present disclosure provide an audio recognition method, relating to the technical field of data processing. The audio recognition method comprises: obtaining second audio obtained by a terminal device internally recording first audio, and third audio obtained by the terminal device externally recording the first audio; determining scene information of the terminal device recording the first audio; determining a first audio feature of the second audio and a second audio feature of the third audio; determining target audio matched with the first audio according to the scene information, the first audio feature and the second audio feature, and determining information of the target audio as recognition information of the first audio. Through the comprehensive use of the audio feature processing of the internally recorded and externally recorded audio and the scene information, a higher quality and consistency audio fingerprint is generated, which greatly improves the recognition rate of the song recognition function and brings a better experience to the user. In addition, the embodiments of the present disclosure provide an audio recognition device, a medium and a computing device.

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relate to the technical field of data processing, and more particularly, to an audio recognition method, device, medium and computing device. BACKGROUND

[0002] This section is intended to provide background or context to the embodiments of the present disclosure. The description herein is not admitted to be prior art merely by inclusion in this section.

[0003] Song recognition refers to when a user hears an unknown audio source, collecting the audio source to identify and return the corresponding song information (including song name, singer, lyrics, etc.) to the user. In related technologies, when identifying, the audio data is often collected to extract the audio fingerprint to identify.

[0004] However, the above implementation is too simple and ignores many unstable factors faced during audio data collection, resulting in inaccurate identification. SUMMARY

[0005] The present disclosure provides an audio recognition method, device, medium and computing device to solve the technical problem of inaccurate song recognition in related technologies.

[0006] In a first aspect of the embodiments of the present disclosure, an audio recognition method is provided, comprising:

[0007] Obtaining second audio obtained by internally recording first audio by a terminal device and third audio obtained by externally recording the first audio by the terminal device;

[0008] Determining scene information of recording the first audio by the terminal device, the scene information including at least one of a device type of the terminal device, noise of an environment where the terminal device is located, and an audio output mode of playing the first audio by the terminal device;

[0009] Determining a first audio feature of the second audio and a second audio feature of the third audio;

[0010] According to the scene information, the first audio feature and the second audio feature, determining a target audio matched with the first audio, and determining information of the target audio as recognition information of the first audio.

[0011] In an embodiment of the present disclosure, the determining a target audio matched with the first audio according to the scene information, the first audio feature and the second audio feature comprises:

[0012] determine a first similarity between the first audio feature and a stored fourth audio feature of an internal recording of the fourth audio, and determine a second similarity between the second audio feature and an external recording audio fingerprint of the fourth audio;

[0013] configure a first weight to the first similarity according to the scene information, and configure a second weight to the second similarity according to the scene information;

[0014] determine a confidence between the first audio and the fourth audio according to the first similarity, the first weight, the second similarity, and the second weight;

[0015] determine the fourth audio as the target audio in response to the confidence being greater than a preset threshold.

[0016] In another embodiment of the present disclosure, the scene information includes the noise and the audio output mode, and the configuring the first weight to the first similarity and the second weight to the second similarity according to the scene information includes:

[0017] determining a noise interval in which the noise is located;

[0018] configuring the first weight to the first similarity and the second weight to the second similarity according to the noise interval and the audio output mode.

[0019] In yet another embodiment of the present disclosure, the configuring the first weight to the first similarity and the second weight to the second similarity according to the noise interval and the audio output mode includes:

[0020] in response to the noise interval indicating that an environment in which the terminal device is located is a low-noise environment and the audio output mode is a speakerphone mode, setting the first weight to be less than the second weight;

[0021] in response to the noise interval indicating that an environment in which the terminal device is located is a low-noise environment and the audio output mode is a headset mode, setting the first weight to be greater than the second weight;

[0022] in response to the noise interval indicating that an environment in which the terminal device is located is a low-noise environment and the audio output mode is a Bluetooth speaker mode, setting the first weight to be equal to the second weight.

[0023] In still another embodiment of the present disclosure, the configuring the first weight to the first similarity and the second weight to the second similarity according to the noise interval and the audio output mode includes:

[0024] in response to the noise range indicating that the terminal device is in a medium-noise environment and the audio output mode being the earphone mode, setting the first weight to be greater than the second weight;

[0025] in response to the noise range indicating that the terminal device is in a medium-noise environment and the audio output mode being the earphone mode, setting the first weight to be greater than the second weight;

[0026] in response to the noise range indicating that the terminal device is in a medium-noise environment and the audio output mode being the earphone mode, setting the first weight to be greater than the second weight.

[0027] In still another embodiment of the present disclosure, the first weight and the second weight are configured according to the noise range and the audio output mode, including:

[0028] in response to the noise range indicating that the terminal device is in a high-noise environment, setting the first weight to be greater than the second weight.

[0029] In still another embodiment of the present disclosure, the first similarity between the first audio feature and the stored fourth audio feature is determined, including:

[0030] obtaining a target parameter, the target parameter including at least one of a current time period and a location where the terminal device is located;

[0031] determining an audio database matching the target parameter, and determining a stored fourth audio in the audio database.

[0032] In still another embodiment of the present disclosure, the first audio feature of the second audio and the second audio feature of the third audio are determined, including:

[0033] obtaining a first proportion of human voice and a second proportion of non-human voice in the first audio;

[0034] determining an audio feature recognition strategy according to the first proportion and the second proportion;

[0035] recognizing the first audio feature of the second audio and the second audio feature of the third audio by using the audio feature recognition strategy.

[0036] In still another embodiment of the present disclosure, the audio feature recognition strategy is determined according to the first proportion and the second proportion, including:

[0037] In response to the first proportion being less than a first preset proportion, or the first proportion being less than a second preset proportion and the second proportion being greater than a third preset proportion, a humming and singing recognition mode is determined as the audio feature recognition strategy.

[0038] In response to the second proportion being less than a third preset proportion, or the second proportion not being less than the third preset proportion and the first proportion not being less than a second preset proportion, the humming and singing recognition mode is determined as the audio feature recognition strategy.

[0039] In still another embodiment of the present disclosure, the employing the audio feature recognition strategy to identify the first audio feature of the second audio and the second audio feature of the third audio comprises:

[0040] In response to the audio feature recognition strategy being the humming and singing recognition mode, after the humming and singing recognition mode is employed to identify the second audio and the third audio for a preset time length, a song lyrics recognition mode is employed to the second audio and the third audio to obtain the first audio feature and the second audio feature.

[0041] In still another embodiment of the present disclosure, the second audio and the third audio have the same starting playback time point.

[0042] In a second aspect of the embodiments of the present disclosure, a computer readable storage medium is provided, and the computer readable storage medium stores computer execution instructions. When a processor executes the computer execution instructions, the method described in the first aspect and any one of the embodiments is implemented.

[0043] In a third aspect of the embodiments of the present disclosure, an audio recognition device is provided, and the audio recognition device comprises:

[0044] An acquisition module is configured to acquire a second audio obtained by internally recording a first audio by a terminal device and a third audio obtained by externally recording the first audio by the terminal device;

[0045] A first determination module is configured to determine scene information of recording the first audio by the terminal device, and the scene information comprises at least one of a device type of the terminal device, noise of an environment where the terminal device is located, and an audio output mode of playing the first audio by the terminal device.

[0046] A second determination module is configured to determine a first audio feature of the second audio and a second audio feature of the third audio.

[0047] determine, according to the scene information, the first audio feature and the second audio feature, a target audio matched with the first audio, and determine information of the target audio as the identification information of the first audio.

[0048] In one embodiment of the present disclosure, the third determining module determines, according to the scene information, the first audio feature and the second audio feature, a target audio matched with the first audio, specifically for:

[0049] determining a first similarity between the first audio feature and a stored fourth audio feature of an internal recording audio, and determining a second similarity between the second audio feature and an external recording audio fingerprint of the fourth audio;

[0050] configuring a first weight for the first similarity according to the scene information, and configuring a second weight for the second similarity according to the scene information;

[0051] determining a confidence between the first audio and the fourth audio according to the first similarity, the first weight, the second similarity and the second weight;

[0052] in response to the confidence being greater than a preset threshold, determining the fourth audio as the target audio.

[0053] In another embodiment of the present disclosure, the scene information includes the noise and the audio output mode, the third determining module configures a first weight for the first similarity according to the scene information, and configures a second weight for the second similarity according to the scene information, specifically for:

[0054] determining a noise interval in which the noise is located;

[0055] configuring the first weight for the first similarity and the second weight for the second similarity according to the noise interval and the audio output mode.

[0056] In yet another embodiment of the present disclosure, the third determining module configures the first weight for the first similarity and the second weight for the second similarity according to the noise interval and the audio output mode, specifically for:

[0057] in response to the noise interval indicating that an environment in which the terminal device is located is a low-noise environment and the audio output mode is an external playing mode, setting the first weight to be less than the second weight;

[0058] in response to the noise interval indicating that an environment in which the terminal device is located is a low-noise environment and the audio output mode is an earphone mode, setting the first weight to be greater than the second weight.

[0059] in response to the noise interval indicating that the terminal device is in a low-noise environment and the audio output mode being a Bluetooth speaker mode, setting the first weight to be equal to the second weight.

[0060] In another embodiment of the present disclosure, the third determining module configures a first weight for the first similarity and a second weight for the second similarity according to the noise interval and the audio output mode, specifically for:

[0061] in response to the noise interval indicating that the terminal device is in a medium-noise environment and the audio output mode being a speakerphone mode, setting the first weight to be equal to the second weight;

[0062] in response to the noise interval indicating that the terminal device is in a medium-noise environment and the audio output mode being a headset mode, setting the first weight to be greater than the second weight;

[0063] in response to the noise interval indicating that the terminal device is in a medium-noise environment and the audio output mode being a Bluetooth speaker mode, setting the first weight to be equal to the second weight.

[0064] In another embodiment of the present disclosure, the third determining module configures a first weight for the first similarity and a second weight for the second similarity according to the noise interval and the audio output mode, specifically for:

[0065] in response to the noise interval indicating that the terminal device is in a high-noise environment, setting the first weight to be greater than the second weight.

[0066] In another embodiment of the present disclosure, the third determining module determines a first similarity between the first audio feature and a stored fourth audio feature of an internal recording audio, specifically for:

[0067] obtaining a target parameter, the target parameter including at least one of a current time period and a location where the terminal device is located;

[0068] determining an audio database matching the target parameter, and determining a stored fourth audio in the audio database.

[0069] In another embodiment of the present disclosure, the second determining module determines a first audio feature of the second audio and a second audio feature of the third audio, specifically for:

[0070] obtaining a first proportion of human voice and a second proportion of non-human voice in the first audio;

[0071] determine an audio feature recognition strategy according to the first proportion and the second proportion;

[0072] identify the first audio feature of the second audio and the second audio feature of the third audio by using the audio feature recognition strategy.

[0073] In still another embodiment of the present disclosure, the second determining module determines an audio feature recognition strategy according to the first proportion and the second proportion, specifically for:

[0074] in response to the first proportion being less than a first preset proportion, or the first proportion being less than a second preset proportion and the second proportion being greater than a third preset proportion, determining a humming and singing recognition mode as the audio feature recognition strategy;

[0075] in response to the second proportion being less than a third preset proportion, or the second proportion not being less than the third preset proportion and the first proportion not being less than a second preset proportion, determining a humming and singing recognition mode as the audio feature recognition strategy.

[0076] In still another embodiment of the present disclosure, the second determining module identifies the first audio feature of the second audio and the second audio feature of the third audio by using the audio feature recognition strategy, specifically for:

[0077] in response to the audio feature recognition strategy being the humming and singing recognition mode, after the second audio and the third audio are identified for a preset time length by using the humming and singing recognition mode, identifying the second audio and the third audio by using a song lyrics recognition mode to obtain the first audio feature and the second audio feature

[0078] In still another embodiment of the present disclosure, the starting playback time point of the second audio and the third audio is the same.

[0079] In a fourth aspect of the embodiments of the present disclosure, a computing device is provided, comprising: at least one processor;

[0080] and a memory connected in communication with the at least one processor;

[0081] wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to cause the computing device to perform the method in the first aspect and any one of the embodiments.

[0082] According to the audio recognition method, device, medium and computing device of the embodiments of the present disclosure, the method comprises: obtaining second audio obtained by internally recording first audio by a terminal device and third audio obtained by externally recording the first audio by the terminal device; determining scene information of recording the first audio by the terminal device; determining first audio features of the second audio and second audio features of the third audio; determining target audio matched with the first audio according to the scene information, the first audio features and the second audio features, and determining information of the target audio as recognition information of the first audio. In the technical solution, by comprehensively using audio feature processing of internally recorded and externally recorded audio and scene information, higher quality and consistency of audio fingerprints are generated, the recognition rate of the song recognition function is greatly improved, and better experience is brought to users. BRIEF DESCRIPTION OF DRAWINGS

[0083] The above and other objects, features and advantages of the present disclosure exemplary embodiments will be more apparent from the following detailed description read in conjunction with the accompanying drawings, in which:

[0084] FIG. 1 A scene schematic diagram of an audio recognition method provided by an embodiment of the present disclosure;

[0085] FIG. 2 A flowchart of an audio recognition method provided by an embodiment of the present disclosure;

[0086] FIG. 3 A flowchart of an audio recognition method provided by another embodiment of the present disclosure;

[0087] FIG. 4 A flowchart of an audio recognition method provided by still another embodiment of the present disclosure;

[0088] FIG. 5 A structure diagram of a storage medium provided by an embodiment of the present disclosure;

[0089] FIG. 6 A structure diagram of an audio recognition device provided by an embodiment of the present disclosure;

[0090] FIG. 7 A structure diagram of a computing device provided by an embodiment of the present disclosure;

[0091] In the drawings, the same or corresponding reference numbers denote the same or corresponding parts. DETAILED DESCRIPTION

[0092] The principles and spirits of the present disclosure will be described below with reference to a number of exemplary embodiments. It should be understood that these embodiments are given only so that those skilled in the art can better understand and implement the present disclosure, and do not limit the scope of the present disclosure in any way. On the contrary, these embodiments are provided so that the present disclosure is more thorough and complete, and the scope of the present disclosure is fully conveyed to those skilled in the art.

[0093] Those skilled in the art understand that the embodiments of the present disclosure can be implemented as a system, device, apparatus, method or computer program product. Therefore, the present disclosure can be embodied in the form of a complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0094] According to embodiments of the present disclosure, a method, medium, device and computing device for audio recognition are provided.

[0095] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present disclosure are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation portal for user to choose authorization or refusal.

[0096] In addition, the number of any elements in the drawings is used for example and not limitation, and any naming is only used for distinction and does not have any limiting meaning.

[0097] The principles and spirits of the present disclosure will be described below with reference to a number of exemplary embodiments. It should be understood that these embodiments are given only so that those skilled in the art can better understand and implement the present disclosure, and do not limit the scope of the present disclosure in any way. On the contrary, these embodiments are provided so that the present disclosure is more thorough and complete, and the scope of the present disclosure is fully conveyed to those skilled in the art. SUMMARY

[0099] The present inventors found that in the related art of song recognition, it is often recognized by collecting audio data and extracting audio fingerprints in the audio data, which has the following technical problems:

[0100] The effect of song recognition is generally limited by many factors, and there will be a situation of reducing the recognition effect, and the simple recognition method in the related art cannot avoid the influence of these factors, thereby resulting in poor recognition effect.

[0101] To solve the technical problems above, the present disclosure provides an audio recognition method, device, medium and computing device. By obtaining audio obtained by internal recording and external recording respectively in a complex use scenario, and performing optimization processing on the audio fingerprint in the obtained audio based on scene information, the accuracy and reliability of audio recognition are significantly improved, and the recognition rate of the song recognition function is greatly improved.

[0102] After introducing the basic principles of the present disclosure, various non-limiting embodiments of the present disclosure will be described in detail.

[0103] Overview of application scenarios

[0104] First, the audio recognition method provided by the embodiments of the present disclosure can be applied to FIG. 1 the scene schematic diagram shown in FIG. 1. FIG. 1 The scene schematic diagram of the audio recognition method provided by the embodiments of the present disclosure. As shown in FIG. 1 , the scene includes a computing device 11 and a terminal device 12.

[0105] In this application scenario, the terminal device 12 collects audio that a user wants to identify by internal recording and external recording respectively, and then the computing device 11 obtains audio corresponding to internal recording and external recording collected by the terminal device 12, and scene information corresponding to the terminal device 12, and obtains an identification result about the identified audio, i.e., a song name, a singer, lyrics, etc., through a corresponding processing process, and then sends the identification result to the terminal device to enable the user to know. The implementation of the processing process can be referred to the schemes of the following embodiments.

[0106] It should be noted that FIG. 1 is only a schematic diagram of an application scenario provided by the embodiments of the present disclosure, and the embodiments of the present disclosure do not limit the devices included in FIG. 1 , nor the positional relationship between the devices in FIG. 1 . For example, in the application scenario shown in FIG. 1 , it can also include related devices corresponding to the song library used by the computing device 11 in matching, and external devices of the terminal device 12, etc.

[0107] Exemplary method

[0108] The method for audio recognition according to the exemplary embodiments of the present disclosure will be described below in combination with the application scenario of FIG. 1 , with reference to FIG. 1 . It should be noted that the above application scenario is only shown for the purpose of facilitating understanding of the spirit and principles of the present disclosure, and the embodiments of the present disclosure are not limited in this respect. On the contrary, the embodiments of the present disclosure can be applied to any applicable scenario.

[0109] FIG. 2 Flowchart of an audio recognition method provided for an embodiment of the present disclosure. The method of the embodiment of the present disclosure can be applied in a computing device, which can be a server. As shown in the figure, the method of the embodiment of the present disclosure comprises: FIG. 2

[0110] S21, obtaining second audio obtained by internally recording the first audio by the terminal device and third audio obtained by externally recording the first audio by the terminal device;

[0111] In this step, there is audio (such as music, cross-talk, etc.) playing on the terminal device, denoted as first audio. At this time, the terminal device internally records the first audio to obtain second audio, and externally records the first audio to obtain third audio.

[0112] Then, the computing device obtains the second audio and the third audio.

[0113] The terminal device can be a mobile phone, a computer, a tablet, etc.

[0114] In a possible implementation, internally recording refers to recording sound using a microphone or an audio input interface inside the terminal device, or refers to sound played in the computer not being played out through a sound box but being recorded directly inside the computer; externally recording refers to recording sound by connecting an external microphone or other audio source to the terminal device, or refers to sound played out through a sound box in the computer being recorded by other recording devices.

[0115] That is, before the implementation of this step, a step that can be performed is that the terminal device detects whether the first audio is currently playing and whether it is in a sound playing out state, and performs recording implementation when both conditions are met.

[0116] Optionally, the second audio and the third audio have the same starting playing time point.

[0117] In this implementation, since there are two different recording modes for identifying the first audio, when recording, in order to make the identification result of the first audio accurate, it is necessary to ensure that the first audio and the second audio have the same starting playing time point, that is, to perform time synchronization processing on the externally recorded third audio and the internally recorded second audio.

[0118] A specific possible implementation can be:

[0119] 1. When starting internal recording and external recording of audio collection, the terminal device records the respective starting time stamps at the same time. When the computing device obtains the above audio data, the recorded time stamps can be used to preliminarily align the audio data corresponding to internal recording and external recording respectively, and the audio signals are corrected synchronously through a cross-correlation method.

[0120] ​2, the terminal device can also compensate for the self-introduced delay based on the recorded timestamp, ensure the starting points of the two audio sources are consistent, and then send to the computing device.

[0121] S22, determine the scene information of the terminal device recording the first audio;

[0122] The scene information includes at least one of the device type of the terminal device, the noise of the environment where the terminal device is located, and the audio output mode of the terminal device playing the first audio.

[0123] In this step, the terminal device can obtain the scene in which the first audio is recorded, and then send the related information of the scene to the computing device, so that the computing device can determine the scene information based on the information.

[0124] Optionally, the scene can include at least one of the following:

[0125] 1, noise, the terminal device can use the microphone to detect the external environmental noise intensity;

[0126] 2, device type, which can include but is not limited to mobile phone, tablet, computer, etc.

[0127] 3, audio output mode, which can include but is not limited to external speaker, earphone, car audio, Bluetooth speaker, etc.

[0128] In addition, there can also be time and geographic location information when the audio is collected.

[0129] That is, in a specific possible implementation, the client corresponding to the song listening and recognition function in the terminal device captures the environmental noise through the microphone and calculates the noise intensity, detects the current terminal device type and audio output mode, and uploads the client to the corresponding server, that is, the computing device of the present disclosure.

[0130] In one possible implementation, the client reports that the noise is low, the device type is a mobile phone, and the audio output mode is external speaker, and the computing device can determine that the scene information for recording the first audio is that the noise is low, the device type is a mobile phone, and the audio output mode is external speaker.

[0131] S23, determine the first audio feature of the second audio and the second audio feature of the third audio;

[0132] In this step, after obtaining the second audio corresponding to the internal recording and the third audio corresponding to the external recording, the audio features of the second audio and the third audio are extracted respectively, that is, the audio fingerprints corresponding to the respective audios can be obtained.

[0133] The audio fingerprint is a digital signature for audio identification, which can uniquely identify a song or an audio segment.

[0134] S24, determine the target audio matched with the first audio according to the scene information, the first audio feature and the second audio feature, and determine the information of the target audio as the recognition information of the first audio.

[0135] In this step, in order to ensure the accuracy of the recognition, after obtaining the first audio feature and the second audio feature, the first audio feature and the second audio feature are processed based on the scene information affecting the recognition accuracy, so as to obtain the target audio matched with the first audio.

[0136] The information of the target audio can be the name, author, download link and the like of the audio.

[0137] In a possible implementation, after matching the audio features of the internal recording and the external recording respectively, two groups of matching results will be obtained. Each group of matching results will include one or more candidate songs and the corresponding confidence. In order to integrate the two groups of results, weights are assigned to the two groups of results according to different use scenarios (device type, audio output mode, ambient noise level), and then weighted processing is performed to obtain the final recognition result. The process in this embodiment can be referred to the embodiment shown in the following FIG. 3 .

[0138] The embodiment of the present disclosure provides an audio recognition method, including: obtaining second audio obtained by internally recording first audio by a terminal device and third audio obtained by externally recording the first audio by the terminal device; determining scene information of recording the first audio by the terminal device; determining a first audio feature of the second audio and a second audio feature of the third audio; determining a target audio matched with the first audio according to the scene information, the first audio feature and the second audio feature, and determining information of the target audio as recognition information of the first audio. By processing the audio features of the internal recording and the external recording audio and comprehensively using the scene information, a higher quality and consistency audio fingerprint is generated, which greatly improves the recognition rate of the song recognition function and brings a better experience to the user.

[0139] Based on the embodiment shown in FIG. 2 , S24 will be further expanded. FIG. 3 The flowchart of the audio recognition method provided by another embodiment of the present disclosure is shown in FIG. 3 , S24 can include the following steps:

[0140] S31, determine a first similarity between the first audio feature and a stored internal recording audio feature of a fourth audio, and determine a second similarity between the second audio feature and an external recording audio fingerprint of the fourth audio;

[0141] In this step, multiple audio features are pre-stored in the computing device, and these audio features can be determined based on an audio database (such as a music library).

[0142] For example, the system can categorize audio data into popular and full libraries based on recent popularity, high-quality and low-quality libraries based on audio quality, and English and Russian libraries based on language. The computing device will allocate server resources reasonably based on the current audio information. For instance, if the collection effect is significantly better between 9 PM and 2 AM than other time periods, or if the collection effect is significantly better on weekends than on weekdays, then the system will prioritize requesting the popular and high-quality libraries during those time periods. Alternatively, it can prioritize requesting the corresponding language library based on the user's geographical location and the identified language information.

[0143] That is, in this embodiment, for the first audio feature, at least one fourth audio with a high similarity to the first audio feature can be identified, and the first similarity can be known; and the similarity between the at least one fourth audio and the second audio feature can be determined, and the second similarity can be known.

[0144] Optionally, the method for determining the first similarity between the first audio feature and the internally recorded audio feature of the stored fourth audio in S31 can be:

[0145] Step 1: Obtain the target parameters, which include at least one of the current time period and the location of the terminal device;

[0146] For example, if the target parameter is the current time period, it can be considered as the preset time period in which the current time period is located. For example, the collection effect is significantly better between 9 pm and 2 am every day than other time periods. For example, the collection effect on weekends is significantly better than on weekdays. In this case, the current time period will prioritize requesting the corresponding audio databases of popular and high-quality libraries.

[0147] For example, if the target parameter is the location of the terminal device, such as the Chinese language usage area, then the audio database corresponding to this target parameter is a Chinese audio database.

[0148] In addition, the usage area can also be a region characterized by local dialects, and there are corresponding audio databases for those local dialects.

[0149] Step 2: Determine the audio database that matches the target parameters, and then determine the fourth audio file stored in the audio database.

[0150] In this implementation, in order to improve the accuracy of recognition, after determining the audio database that matches the target parameters, the fourth audio stored in the audio database is determined.

[0151] S32. Configure a first weight for the first similarity based on the scene information, and configure a second weight for the second similarity based on the scene information;

[0152] In this step, based on the received scene information such as environmental noise information, device type and audio output mode information, a set weight distribution strategy can be used to assign corresponding weights to the first and second similarities.

[0153] Optionally, the scene information includes noise and audio output mode, and then:

[0154] In a possible implementation, S32 can be implemented by the following steps:

[0155] Step 1, determine the noise interval in which the noise is located;

[0156] In this implementation, the noise interval in which the noise is located can be determined based on the intensity of the noise.

[0157] The noise interval can be low noise (e.g., less than 40 dB), medium noise (e.g., between 40 dB and 70 dB), and high noise (e.g., greater than 70 dB).

[0158] Step 2, configure the first weight for the first similarity and the second weight for the second similarity according to the noise interval and the audio output mode.

[0159] In the example, the noise interval can be divided into low noise, medium noise, and high noise, and for different noise intervals, the following possible implementations exist respectively:

[0160] First, when the noise interval indicates that the terminal device is located in a low noise environment:

[0161] If the audio output mode is an external playback mode, the first weight is set to be less than the second weight; if the audio output mode is an earphone mode, the first weight is set to be greater than the second weight; and if the audio output mode is a Bluetooth speaker mode, the first weight is set to be equal to the second weight.

[0162] For example, in the external playback mode, the external recording audio fingerprint weight is high and the internal recording audio fingerprint weight is low; in the earphone mode, the internal recording audio fingerprint weight is high and the external recording audio fingerprint weight is low; and in the Bluetooth speaker mode, the external recording audio fingerprint weight and the internal recording audio fingerprint weight are equivalent.

[0163] Second, when the noise interval indicates that the terminal device is located in a medium noise environment:

[0164] If the audio output mode is an external playback mode, the first weight is set to be equal to the second weight; if the audio output mode is an earphone mode, the first weight is set to be greater than the second weight; and if the audio output mode is a Bluetooth speaker mode, the first weight is set to be equal to the second weight.

[0165] For example, in the outdoor mode, the outdoor recording audio fingerprint and the indoor recording audio fingerprint have equal weights; in the earphone mode, the indoor recording audio fingerprint has a higher weight, and the outdoor recording audio fingerprint has a lower weight; and in the Bluetooth speaker mode, the outdoor recording audio fingerprint and the indoor recording audio fingerprint have equal weights.

[0166] The third: when the noise interval indicates that the terminal device is in a high-noise environment:

[0167] The audio output mode is the outdoor mode, the first weight is set to be less than the second weight; the audio output mode is the earphone mode, the first weight is set to be greater than the second weight; and the audio output mode is the Bluetooth speaker mode, the first weight is set to be greater than the second weight.

[0168] For example, in the outdoor mode, the outdoor recording audio fingerprint is not less than the indoor recording audio fingerprint; in the earphone mode, the indoor recording audio fingerprint has a higher weight, and the outdoor recording audio fingerprint has a lower weight; and in the Bluetooth speaker mode, the outdoor recording audio fingerprint and the indoor recording audio fingerprint have equal weights.

[0169] S33, determining a confidence between the first audio and the fourth audio according to the first similarity, the first weight, the second similarity, and the second weight;

[0170] In this implementation, taking the fourth audio as the song A, the confidence of the song A can be the first similarity*the first weight+the second similarity*the second weight.

[0171] For example, the first similarity is 0.85, the first weight is 0.3, the second similarity is 0.80, and the second weight is 0.7, and the confidence between the first audio and the song A is:

[0172] 0.85*0.3+0.80*0.7=0.815

[0173] S34, determining the fourth audio as the target audio in response to the confidence being greater than a preset threshold.

[0174] In this step, in order to ensure the accuracy of the recommendation, a preset threshold needs to be set to filter the confidence between the possible first audio and the fourth audio, and the fourth audio with a greater confidence can be considered as the target audio.

[0175] For example, the preset threshold is 0.2.

[0176] It should be understood that there can be at least one fourth audio, and in the subsequent output process, at least the first fourth audio with a confidence greater than the preset threshold can be determined as the target audio, and the target audio can be fed back to the client and displayed in order according to the size of the confidence.

[0177] For example, the confidence of song A is 0.815, the confidence of song B is 0.49, and the confidence of song C is 0.225. According to the order from high to low of the confidence, the song list [A, C, B] is returned to the client and presented to the user.

[0178] An embodiment of the present disclosure provides an audio recognition method, comprising: determining a first similarity between a first audio feature and a stored fourth audio's internal recording audio feature, and determining a second similarity between a second audio feature and a fourth audio's external recording audio fingerprint; configuring a first weight for the first similarity according to scene information, and configuring a second weight for the second similarity according to scene information; determining a confidence between the first audio and the fourth audio according to the first similarity, the first weight, the second similarity, and the second weight; and determining the fourth audio as a target audio in response to the confidence being greater than a preset threshold. In the technical solution, after the internal recording and external recording audio fingerprints are matched respectively, two matching results will be obtained, and each matching result will include one or more candidate songs and their corresponding confidence. In order to integrate the two sets of results, weights are assigned to the two results according to different use scenarios (device type, audio output mode, and environmental noise level), and then weighted processing is performed, which can greatly improve the accuracy of recognition.

[0179] Based on the above embodiment, S23 is further expanded. FIG. 4 The flowchart of the audio recognition method provided by another embodiment of the present disclosure is shown in FIG. 4. FIG. 4 As shown in FIG. 4, S23 can include the following steps:

[0180] S41, obtaining a first proportion of human voice and a second proportion of non-human voice in the first audio;

[0181] After this step, there are several different recognition modes according to the size of the proportion of non-human voice (such as the proportion of music) and the proportion of human voice. Therefore, in this step, the first proportion of human voice and the second proportion of non-human voice in the first audio are first obtained.

[0182] In one possible implementation, the average volume or loudness of human voice and non-human voice can be measured first, and then the proportion of human voice and non-human voice is calculated according to the measured volume data, that is, the first proportion of human voice and the second proportion of non-human voice are obtained.

[0183] It should be understood that, since the volume of human voice and non-human voice may change dynamically, the measurement results at multiple time points can be considered for comprehensive processing.

[0184] S42, determining an audio feature recognition strategy according to the first proportion and the second proportion;

[0185] In this step, different audio feature recognition strategies can be adopted according to the difference between the first proportion and the second proportion of non-human voice, for example, can include: song recognition mode, humming recognition mode, and lyrics recognition mode (see the embodiment in S43).

[0186] Optionally, S42 can have the following possible implementations:

[0187] 1st, in response to the first proportion being less than the first preset proportion, or the first proportion being less than the second preset proportion and the second proportion being greater than the third preset proportion, the song recognition recognition mode is determined as the audio feature recognition strategy;

[0188] In this implementation, for example, the first preset proportion can be set to 6%, the second preset proportion can be set to 70%, and the third preset proportion can be set to 3%;

[0189] That is, the proportion of human voice is analyzed first, when the proportion of human voice is less than 6%, the song recognition mode is used; secondly, the proportion of music is analyzed, when the proportion of music is higher than 3%, and when the proportion of human voice does not exceed 70%, the song recognition recognition mode is used.

[0190] 2nd, in response to the second proportion being less than the third preset proportion, or the second proportion not being less than the third preset proportion and the first proportion not being less than the second preset proportion, the humming recognition recognition mode is determined as the audio feature recognition strategy;

[0191] That is, the proportion of music is analyzed, when the proportion of music is less than 3%, the humming recognition mode is used first; when the proportion of music is higher than 3%, and the proportion of human voice exceeds 70%, the humming recognition recognition mode is used first.

[0192] S43, using the audio feature recognition strategy, identifying the first audio feature of the second audio and the second audio feature of the third audio.

[0193] In this step, after the audio feature recognition strategy is determined above, the first audio feature and the second audio feature obtained above are identified based on the corresponding recognition mode.

[0194] That is, this implementation mainly extracts representative features from each frame of audio signal, and constructs an audio fingerprint with uniqueness and stability based on the extracted features, that is, as an audio feature, the audio fingerprint is a digital signature for audio recognition, which can uniquely identify a song or an audio segment.

[0195] Optionally, S43 can have the following possible implementations:

[0196] In response to the audio feature recognition strategy being the humming recognition mode, the humming recognition mode is used to recognize the second audio and the third audio for a preset time length, and then the lyric recognition mode is used to recognize the second audio and the third audio to obtain the first audio feature and the second audio feature.

[0197] In this implementation, for example, the preset time length can be set to 10 seconds.

[0198] That is, when the humming mode does not have a result at the 10th second, the lyric recognition mode is used to recognize the second audio and the third audio to obtain the first audio feature and the second audio feature.

[0199] The embodiment of the present disclosure provides an audio recognition method, comprising: obtaining a first proportion of human voice and a second proportion of non-human voice in first audio; determining an audio feature recognition strategy according to the first proportion and the second proportion; and using the audio feature recognition strategy to recognize a first audio feature of second audio and a second audio feature of third audio. In this technical solution, different recognition modes are configured for the proportions of non-human voice and human voice to configure different recognition strategies, and the accuracy of recognition is improved when the recognition direction is refined.

[0200] Exemplary medium

[0201] After introducing the method of the exemplary embodiment of the present disclosure, next, reference is made to FIG. 5 The storage medium of the exemplary embodiment of the present disclosure is described.

[0202] Reference is made to FIG. 5 As shown in the figure, the storage medium 50 stores a program product for implementing the above method according to the embodiment of the present disclosure, which can adopt a portable compact disc read-only memory (CD-ROM) and include program codes, and can run on a computing device, such as a server. However, the program product of the present disclosure is not limited to this.

[0203] The program product can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium may, for example, be but is not limited to an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0204] A readable signal medium can include a data signal traveling in baseband or traveling as a part of a carrier wave traveling in baseband, wherein the data signal carries the readable program code. Such a data signal can be of any form and can travel over any communication channel, such as metal cables, fiber optics, or wireless channels, including any suitable combination of the above. A readable signal medium can also be any medium that can be read by a machine.

[0205] The program code for carrying out operations of the present disclosure can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++, or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's device, partly on a remote computing device, or entirely on the remote computing device or server. In the latter scenario, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computing device, such as through the Internet using an Internet Service Provider (ISP).

[0206] Exemplary apparatus

[0207] After introducing the medium of the exemplary embodiments of the present disclosure, next, the audio recognition device of the exemplary embodiments of the present disclosure is described with reference to the drawings, which is used to implement the method in any of the above method embodiments, and the implementation principle and technical effects are similar, which will not be described here.

[0208] FIG. 6 The structural diagram of the audio recognition device provided by an embodiment of the present disclosure is shown in FIG. 1. As shown in FIG. 1, the audio recognition device includes: FIG. 6

[0209] The acquisition module 61 is configured to acquire second audio obtained by the terminal device internally recording the first audio and third audio obtained by the terminal device externally recording the first audio.

[0210] The first determination module 62 is configured to determine scene information of the terminal device recording the first audio, and the scene information includes at least one of a device type of the terminal device, noise of an environment where the terminal device is located, and an audio output mode of the terminal device playing the first audio.

[0211] The second determination module 63 is configured to determine a first audio feature of the second audio and a second audio feature of the third audio.

[0212] The third determination module 64 is configured to determine a target audio matched with the first audio according to the scene information, the first audio feature, and the second audio feature, and determine information of the target audio as the recognition information of the first audio.

[0213] ​In one embodiment of the present disclosure, the third determining module 64 determines the target audio matched with the first audio according to the scene information, the first audio feature, and the second audio feature, specifically for:

[0214] determining a first similarity between the first audio feature and the stored fourth audio's internal recording audio feature, and determining a second similarity between the second audio feature and the fourth audio's external recording audio fingerprint;

[0215] configuring a first weight to the first similarity according to the scene information, and configuring a second weight to the second similarity according to the scene information;

[0216] determining a confidence between the first audio and the fourth audio according to the first similarity, the first weight, the second similarity, and the second weight;

[0217] determining the fourth audio as the target audio in response to the confidence being greater than a preset threshold.

[0218] In another embodiment of the present disclosure, the scene information includes noise and audio output mode, the third determining module 64 configures a first weight to the first similarity according to the scene information, and configures a second weight to the second similarity according to the scene information, specifically for:

[0219] determining a noise interval in which the noise is located;

[0220] configuring the first weight to the first similarity and the second weight to the second similarity according to the noise interval and the audio output mode.

[0221] In yet another embodiment of the present disclosure, the third determining module 64 configures the first weight to the first similarity and the second weight to the second similarity according to the noise interval and the audio output mode, specifically for:

[0222] in response to the noise interval indicating that the terminal device is located in a low-noise environment, and the audio output mode being an external playing mode, setting the first weight to be less than the second weight;

[0223] in response to the noise interval indicating that the terminal device is located in a low-noise environment, and the audio output mode being an earphone mode, setting the first weight to be greater than the second weight;

[0224] in response to the noise interval indicating that the terminal device is located in a low-noise environment, and the audio output mode being a Bluetooth speaker mode, setting the first weight to be equal to the second weight.

[0225] In still another embodiment of the present disclosure, the third determining module 64 configures the first weight to the first similarity and the second weight to the second similarity according to the noise interval and the audio output mode, specifically for:

[0226] In response to the noise interval indicating that the terminal device is located in a medium-noise environment and the audio output mode being the speaker mode, the first weight is set to be equal to the second weight;

[0227] In response to the noise interval indicating that the terminal device is located in a medium-noise environment and the audio output mode being the earphone mode, the first weight is set to be greater than the second weight;

[0228] In response to the noise interval indicating that the terminal device is located in a medium-noise environment and the audio output mode being the Bluetooth speaker mode, the first weight is set to be equal to the second weight.

[0229] In still another embodiment of the present disclosure, the third determination module 64 configures a first weight for the first similarity and a second weight for the second similarity according to the noise interval and the audio output mode, and is specifically configured to:

[0230] In response to the noise interval indicating that the terminal device is located in a high-noise environment, the first weight is set to be greater than the second weight.

[0231] In still another embodiment of the present disclosure, the third determination module 64 determines the first similarity between the first audio feature and the stored fourth audio feature of the recorded audio, and is specifically configured to:

[0232] Obtain a target parameter, the target parameter including at least one of a current time period and a location where the terminal device is located;

[0233] Determine an audio database matching the target parameter, and determine the stored fourth audio in the audio database.

[0234] In still another embodiment of the present disclosure, the second determination module 63 determines the first audio feature of the second audio and the second audio feature of the third audio, and is specifically configured to:

[0235] Obtain a first proportion of human voice and a second proportion of non-human voice in the first audio;

[0236] Determine an audio feature recognition strategy according to the first proportion and the second proportion;

[0237] Identify the first audio feature of the second audio and the second audio feature of the third audio by using the audio feature recognition strategy.

[0238] In still another embodiment of the present disclosure, the second determination module 63 determines the audio feature recognition strategy according to the first proportion and the second proportion, and is specifically configured to:

[0239] In response to the first proportion being less than a first preset proportion, or the first proportion being less than a second preset proportion and the second proportion being greater than a third preset proportion, the recognition mode of the song recognition is determined as the audio feature recognition strategy;

[0240] In response to the second proportion being less than the third preset proportion, or the second proportion being not less than the third preset proportion and the first proportion being not less than the second preset proportion, the humming melody recognition mode is determined as the audio feature recognition strategy.

[0241] In still another embodiment of the present disclosure, the second determination module 63 adopts the audio feature recognition strategy to identify the first audio feature of the second audio and the second audio feature of the third audio, specifically for:

[0242] In response to the audio feature recognition strategy being the humming melody recognition mode, after the second audio and the third audio are identified for a preset time length by using the humming melody recognition mode, the lyrics melody recognition mode is used to identify the second audio and the third audio to obtain the first audio feature and the second audio feature.

[0243] In still another embodiment of the present disclosure, the starting playback time points of the second audio and the third audio are the same.

[0244] The audio recognition apparatus provided by the embodiments of the present disclosure can be used to execute the audio recognition method in any of the above embodiments, and has similar implementation principles and technical effects, which will not be described here again.

[0245] Exemplary computing device

[0246] After the method, medium and apparatus of the exemplary embodiments of the present disclosure are introduced, next, reference is made to FIG. 7 The computing device of the exemplary embodiments of the present disclosure is described.

[0247] FIG. 7 The displayed computing device (which can be the computing device 11 described above) is merely an example, and should not bring any limitation to the functions and use range of the embodiments of the present disclosure.

[0248] As shown in FIG. 7 The computing device is in the form of a general-purpose computing device. The components of the computing device can include, but are not limited to, at least one processing unit 71, at least one storage unit 72, and a bus 73 connecting different system components (including the processing unit 71 and the storage unit 72). The at least one storage unit 72 stores computer execution instructions; the at least one processing unit 71 includes a processor, which executes the computer execution instructions to implement the method described above.

[0249] The bus 73 includes a data bus, a control bus and an address bus.

[0250] The storage unit 72 can include a readable medium that can be non- transitory in nature and that can include a volatile or non-volatile memory, such as a RAM 721, and / or a cache 722. The storage unit 72 can also include a read only memory (ROM) 723.

[0251] The storage unit 72 can also include the program / utility 725 having a set of program modules 727 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which can generally be referred to as being included in the program components 725. These components can be executed by the processor 710 as a result of being stored in the working memory unit 712 or other volatile memory device.

[0252] The computing device can also communicate with one or more external devices 77 such as a keyboard or a pointing device, through an I / O interface 75. The I / O interface 75 can also include a communication device such as a modem, a network card, or an infrared communication device, to enable the computing device to communicate with other devices and systems. The communication device can enable the computing device to communicate with other devices and systems through a network 78, such as a local area network (LAN), a wide area network (WAN), and / or the Internet, for example. FIG. 7 As shown, the network adapter 76 communicates with the other components of the computing device through the bus 73. It should be understood that although not shown, other hardware and / or software components that are coupled to the computing device can provide the necessary interfaces to enable communication between the components and the networks 78.

[0253] It should be noted that although several units / modules or sub-units / modules of the audio recognition apparatus are mentioned in the foregoing detailed description, such division is merely exemplary and not mandatory. Indeed, according to embodiments of the present disclosure, features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, features and functions of one unit / module described above can be further divided into units / modules.

[0254] Moreover, while operations of the methods of the present disclosure are described in a particular order in the figures, this is not required or implied in any particular order for performing the operations to achieve desirable results. Additionally or alternatively, certain steps can be omitted, combined, performed in a different order, and / or split into multiple steps.

[0255] While the spirit and principles of the present disclosure have been described with reference to several specific embodiments, it is to be understood that the present disclosure is not limited to the specific embodiments disclosed and that the division of the aspects is not meant to imply that features from the aspects cannot be combined to benefit from the disclosure, but is merely for convenience of presentation. The present disclosure is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the claims appended hereto.

Claims

1. An audio recognition method, characterized in that, include: The terminal device acquires a second audio recording obtained by internally recording a first audio and a third audio recording obtained by externally recording the first audio. Determine the scene information in which the terminal device records the first audio, the scene information including at least one of the following: the device type of the terminal device, the noise of the environment in which the terminal device is located, and the audio output mode in which the terminal device plays the first audio. Determine the first audio feature of the second audio and the second audio feature of the third audio; Based on the scene information, the first audio feature, and the second audio feature, a target audio matching the first audio is determined, and the information of the target audio is determined as the recognition information of the first audio.

2. The audio recognition method according to claim 1, characterized in that, The step of determining the target audio matching the first audio based on the scene information, the first audio feature, and the second audio feature includes: Determine the first similarity between the first audio feature and the internally recorded audio feature of the stored fourth audio, and determine the second similarity between the second audio feature and the externally recorded audio fingerprint of the fourth audio; A first weight is configured for the first similarity based on the scene information, and a second weight is configured for the second similarity based on the scene information; The confidence level between the first audio and the fourth audio is determined based on the first similarity, the first weight, the second similarity, and the second weight. In response to the confidence level being greater than a preset threshold, the fourth audio is determined as the target audio.

3. The audio recognition method according to claim 2, characterized in that, The scene information includes the noise and the audio output mode. The step of configuring a first weight for the first similarity based on the scene information, and configuring a second weight for the second similarity based on the scene information, includes: Determine the noise range in which the noise is located; Based on the noise range and the audio output mode, a first weight is configured for the first similarity and a second weight is configured for the second similarity.

4. The audio recognition method according to claim 3, characterized in that, The step of configuring a first weight for the first similarity and a second weight for the second similarity based on the noise range and the audio output mode includes: In response to the noise range indicating that the terminal device is in a low-noise environment and the audio output mode is external speaker mode, the first weight is set to be less than the second weight. In response to the noise range indicating that the terminal device is in a low-noise environment and the audio output mode is headphone mode, the first weight is set to be greater than the second weight. In response to the noise range indicating that the terminal device is in a low-noise environment and the audio output mode is Bluetooth speaker mode, the first weight is set to be equal to the second weight.

5. The audio recognition method according to claim 3, characterized in that, The step of configuring a first weight for the first similarity and a second weight for the second similarity based on the noise range and the audio output mode includes: In response to the noise range indicating that the terminal device is in a medium noise environment and the audio output mode is external speaker mode, the first weight is set to be equal to the second weight; In response to the noise range indicating that the terminal device is in a medium noise environment and the audio output mode is headphone mode, the first weight is set to be greater than the second weight; In response to the noise range indicating that the terminal device is in a medium noise environment and the audio output mode is Bluetooth speaker mode, the first weight is set to be equal to the second weight.

6. The audio recognition method according to claim 3, characterized in that, The step of configuring a first weight for the first similarity and a second weight for the second similarity based on the noise range and the audio output mode includes: In response to the noise range indicating that the terminal device is in a high-noise environment, the first weight is set to be greater than the second weight.

7. The audio recognition method according to claim 2, characterized in that, Determining the first similarity between the first audio feature and the internally recorded audio features of the stored fourth audio includes: Obtain target parameters, wherein the target parameters include at least one of the current time period and the location of the terminal device; Identify an audio database that matches the target parameters, and determine the fourth audio stored in the audio database.

8. The audio recognition method according to claim 1, characterized in that, Determining the first audio feature of the second audio and the second audio feature of the third audio includes: Obtain the first proportion of human voices and the second proportion of non-human voices in the first audio; Based on the first proportion and the second proportion, an audio feature recognition strategy is determined; The audio feature recognition strategy is used to identify the first audio feature of the second audio and the second audio feature of the third audio.

9. The audio recognition method according to claim 8, characterized in that, The step of determining the audio feature recognition strategy based on the first proportion and the second proportion includes: In response to the first proportion being less than the first preset proportion, or the first proportion being less than the second preset proportion and the second proportion being greater than the third preset proportion, the recognition mode of song recognition is determined as the audio feature recognition strategy. In response to the second proportion being less than the third preset proportion, or the second proportion being not less than the third preset proportion and the first proportion being not less than the second preset proportion, the humming recognition mode is determined as the audio feature recognition strategy.

10. The audio recognition method according to claim 8, characterized in that, The step of using the audio feature recognition strategy to identify the first audio feature of the second audio and the second audio feature of the third audio includes: In response to the audio feature recognition strategy being a humming recognition mode, the second audio and the third audio are recognized for a preset duration using the humming recognition mode, and then the second audio and the third audio are recognized using a lyrics recognition mode to obtain the first audio feature and the second audio feature.

11. The audio recognition method according to any one of claims 1-10, characterized in that, The second audio and the third audio have the same start playback time.

12. An audio recognition device, characterized in that, include: The acquisition module is used to acquire a second audio obtained by the terminal device internally recording the first audio and a third audio obtained by the terminal device externally recording the first audio. The first determining module is used to determine the scene information of the terminal device recording the first audio, the scene information including at least one of the device type of the terminal device, the noise of the environment in which the terminal device is located, and the audio output mode of the terminal device playing the first audio. The second determining module is used to determine the first audio feature of the second audio and the second audio feature of the third audio. The third determining module is used to determine a target audio that matches the first audio based on the scene information, the first audio feature, and the second audio feature, and to determine the information of the target audio as the recognition information of the first audio.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, implement the method as described in any one of claims 1 to 11.

14. A computing device, characterized in that, include: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, cause the computing device to perform the method as described in any one of claims 1 to 11.

Images