Speech recognition method and device, electronic equipment and storage medium

Abstract

According to the speech recognition method and device, the electronic equipment and the storage medium provided by the invention, the to-be-recognized speech collected by the target speech collection equipment is acquired, the audio extraction processing is performed on the to-be-recognized speech, and multiple audios corresponding to different users are separated from the to-be-recognized speech; selecting an audio in a target voice area corresponding to the target voice collection equipment from the plurality of audios as a target audio; and performing voiceprint recognition on the target audio, determining a target user identity corresponding to the target audio according to a voiceprint recognition result, and executing a voice operation corresponding to the target audio by using the target user identity under the condition that the target user identity is a registered user. Compared with the prior art, the embodiment of the invention has the advantages that the corresponding target audio is determined through audio extraction and voice region separation, and the target user identity corresponding to the target audio is determined through voiceprint recognition, so that the user identity of the voice for interaction can be accurately confirmed, and voice interaction can be accurately performed in a multi-person voice interaction scene.

Description

Technical Field

[0001] The present disclosure relates to the technical field of voice processing, and in particular, to a method and apparatus for speech recognition, an electronic device, and a storage medium. Background Art

[0002] With the development of technology, speech recognition technology has been applied in various fields and entered our daily lives. However, in the current common speech recognition application scenarios, generally, a single user interacts with a speech recognition application in an independent physical space. In this case, the speech recognition application directly recognizes the speech of the single user collected by the microphone, obtains the semantics of the speech of the single user, and then interacts with the user according to the semantics.

[0003] However, for the scenario where multiple people interact with a speech recognition application simultaneously in the same physical space, since the microphone will simultaneously collect the voices of other people in the same physical space, the speech recognition application will not be able to confirm which voice in the microphone to interact with, resulting in a failed interaction. For example, in the same game hall, if multiple players play game interactions in the same voice-controlled game, when two or more players simultaneously issue voice operation instructions through their respective microphones, there will be a situation where the microphones of the players simultaneously collect the voice operation instructions issued by other players. The voice-controlled game will not be able to recognize which voice in the microphone is the player using the microphone. For example, if two voices "go forward" and "go backward" are collected in a microphone, the voice-controlled game will not be able to determine which game character to control and which voice operation instruction to execute, resulting in the game being unable to proceed. Summary of the Invention

[0004] The present disclosure provides a method and apparatus for speech recognition, an electronic device, and a storage medium. Its main purpose is to solve the problem that in the scenario where multiple people interact with a speech recognition application simultaneously in the same physical space, since the microphone will simultaneously collect the voices of other people in the same physical space, the speech recognition application will not be able to confirm which voice in the microphone to interact with, resulting in a failed interaction.

[0005] According to a first aspect of the present disclosure, there is provided a method for speech recognition, which includes:

[0006] Obtain the speech to be recognized collected by a target voice collection device, perform audio extraction processing on the speech to be recognized, and separate multiple audio corresponding to different users from the speech to be recognized. The speech to be recognized contains audio corresponding to different users, and the audio corresponding to different users is different;

[0007] From the multiple audio files, select the audio file that is within the target audio region corresponding to the target voice collection device as the target audio file. The target audio region is a pre-divided physical space corresponding to the target voice collection device. Perform voiceprint recognition on the target audio file, and determine the identity of the target user corresponding to the target audio file according to the result of the voiceprint recognition. In the case where the target user identity is a registered user, execute the voice operation corresponding to the target audio file using the target user identity.

[0008] Optionally, the step of selecting the audio file that is within the target audio region corresponding to the target voice collection device as the target audio file from the multiple audio files includes:

[0009] Determine the sound parameters corresponding to each audio file among the multiple audio files. The sound parameters are at least used to characterize the volume of the corresponding audio file.

[0010] Based on the sound parameters and the location information of the target voice collection device, locate the sound source of each audio file, and determine the spatial position of the sound source corresponding to each audio file.

[0011] Determine the audio file whose sound source spatial position is within the target audio region as the target audio file.

[0012] Optionally, the step of determining the audio file whose sound source spatial position is within the target audio region as the target audio file includes:

[0013] If there is an overlapping region between the target audio region and other audio regions of other voice collection devices, then when it is detected that the spatial position of the sound source is within the overlapping region, calculate the distances between the sound source and the target voice collection device and other voice collection devices respectively.

[0014] When the distance between the sound source within the overlapping region and the target voice collection device is the shortest, determine the audio file corresponding to the sound source as the target audio file.

[0015] Optionally, before performing voiceprint recognition on the target audio file, it further includes:

[0016] Perform semantic recognition on the target audio file to obtain the semantic recognition result corresponding to the target audio file.

[0017] Match the semantic recognition result in a preset instruction library.

[0018] If a corresponding voice control instruction is matched in the preset instruction library, then extract the voiceprint features of the target audio file to obtain the corresponding target voiceprint information, and perform voiceprint recognition on the target audio file according to the target voiceprint information.

[0019] If no corresponding voice control instruction is matched in the preset instruction library, determine that the target audio is invalid audio and delete the invalid audio.

[0020] Optionally, determining the target user identity corresponding to the target audio according to the result of voiceprint recognition includes:

[0021] Match the target voiceprint information in the preset voiceprint database;

[0022] If the voiceprint corresponding to the target voiceprint information is matched, determine that the target user identity corresponding to the target audio is a registered user;

[0023] If the voiceprint corresponding to the target voiceprint information is not matched, determine that the target user identity corresponding to the target audio is an unregistered user.

[0024] Optionally, before performing audio extraction processing on the speech to be recognized, it further includes:

[0025] Detect whether there is a sound within a preset human voice frequency spectrum range in the speech to be recognized, and the preset human voice frequency spectrum range is the frequency spectrum range of the user's voice;

[0026] If there is a sound within the preset human voice frequency spectrum range, and it is determined that there is also a sound not within the preset human voice frequency spectrum range in the speech to be recognized, determine the sound not within the preset human voice frequency spectrum range as noise, eliminate the noise in the speech to be recognized, and perform audio extraction processing on the speech to be recognized after eliminating the noise;

[0027] If there is a sound within the preset human voice frequency spectrum range, and it is determined that there is no sound not within the preset human voice frequency spectrum range in the speech to be recognized, directly perform audio extraction processing on the speech to be recognized;

[0028] If there is no sound within the preset human voice frequency spectrum range, determine that the sound in the speech to be recognized is noise and delete the speech to be recognized.

[0029] Optionally, the method further includes:

[0030] In the case where the target audio contains at least one target audio corresponding to a registered user and a target audio corresponding to an unregistered user, preferentially execute the voice operation corresponding to the target audio corresponding to the registered user.

[0031] According to a second aspect of the present disclosure, there is provided a voice recognition device, including:

[0032] An extraction unit that obtains the speech to be recognized collected by a target speech collection device, performs audio extraction processing on the speech to be recognized, and separates multiple audio corresponding to different users from the speech to be recognized. The speech to be recognized contains audio corresponding to different users, and the audio corresponding to different users is different;

[0033] A selection unit for selecting, from the multiple audio, the audio information in the target sound area corresponding to the target speech collection device as the target audio. The target sound area is a pre-divided physical space corresponding to the target speech collection device;

[0034] An identification unit for performing voiceprint recognition on the target audio and determining the target user identity corresponding to the target audio according to the result of the voiceprint recognition;

[0035] An operation unit for, when the target user identity is a registered user, performing the voice operation corresponding to the target audio in the target user identity.

[0036] Optionally, the selection unit is further configured to:

[0037] Determine the sound parameters corresponding to each audio in the multiple audio, where the sound parameters are at least used to characterize the volume size of the corresponding audio;

[0038] Based on the sound parameters and the position information of the target speech collection device, locate the sound source of each audio and determine the spatial position of the sound source corresponding to each audio;

[0039] Determine the audio whose sound source spatial position is in the target sound area as the target audio.

[0040] Optionally, the selection unit is further configured to:

[0041] When there is an overlapping area between the target sound area and other sound areas of other speech collection devices, in the case of detecting that the spatial position of the sound source is in the overlapping area, calculate the distances between the sound source and the target speech collection device and other speech collection devices respectively;

[0042] When the distance between the sound source in the overlapping area and the target speech collection device is the shortest, determine the audio corresponding to the sound source as the target audio.

[0043] Optionally, the device further includes:

[0044] The identification unit is further configured to perform semantic recognition on the target audio to obtain a semantic recognition result corresponding to the target audio;

[0045] A matching unit for matching the semantic recognition result in a preset instruction library;

[0046] The recognition unit is further configured to, when a corresponding voice control instruction is matched in the preset instruction library, extract a voiceprint feature from the target audio to obtain corresponding target voiceprint information, and perform voiceprint recognition on the target audio according to the target voiceprint information;

[0047] The recognition unit is further configured to, when no corresponding voice control instruction is matched in the preset instruction library, determine that the target audio is an invalid audio, and delete the invalid audio.

[0048] Optionally, the recognition unit includes:

[0049] A matching module for matching the target voiceprint information in a preset voiceprint database;

[0050] A determination module for determining that the target user identity corresponding to the target audio is a registered user when the voiceprint corresponding to the target voiceprint information is matched;

[0051] The determination module is further configured to determine that the target user identity corresponding to the target audio is an unregistered user when the voiceprint corresponding to the target voiceprint information is not matched.

[0052] Optionally, the device further includes:

[0053] A detection unit for detecting whether there is a sound within a preset human voice spectrum range in the voice to be recognized, where the preset human voice spectrum range is the spectrum range of the user's voice;

[0054] A clearing unit for, when there is a sound within the preset human voice spectrum range and it is determined that there is also a sound not within the preset human voice spectrum range in the voice to be recognized, determining the sound not within the preset human voice spectrum range as noise, eliminating the noise in the voice to be recognized, and performing audio extraction processing on the voice to be recognized after noise elimination;

[0055] The extraction unit is further configured to, when there is a sound within the preset human voice spectrum range and it is determined that there is no sound not within the preset human voice spectrum range in the voice to be recognized, directly perform audio extraction processing on the voice to be recognized;

[0056] A deletion unit for, when there is no sound within the preset human voice spectrum range, determining the sound in the voice to be recognized as noise and deleting the voice to be recognized.

[0057] Optionally, when the target audio includes at least one target audio corresponding to a registered user and a target audio corresponding to an unregistered user, the operation unit is further configured to preferentially execute the voice operation corresponding to the target audio corresponding to the registered user.

[0058] According to a third aspect of the present disclosure, there is provided an electronic device, including:

[0059] at least one processor; and

[0060] a memory communicatively connected to the at least one processor; wherein,

[0061] the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method described in the foregoing first aspect.

[0062] According to a fourth aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the method described in the foregoing first aspect.

[0063] According to a fifth aspect of the present disclosure, there is provided a computer program product including a computer program, and the computer program implements the method described in the foregoing first aspect when executed by a processor.

[0064] The method and apparatus for speech recognition, electronic device, and storage medium provided by the present disclosure obtain the speech to be recognized collected by the target speech collection device, perform audio extraction processing on the speech to be recognized, and separate multiple audios corresponding to different users from the speech to be recognized. The speech to be recognized contains audios corresponding to different users, and the audios corresponding to different users are different. From the multiple audios, select the audio in the target sound area corresponding to the target speech collection device as the target audio, where the target sound area is a pre-divided physical space corresponding to the target speech collection device. Perform voiceprint recognition on the target audio, and determine the target user identity corresponding to the target audio according to the result of the voiceprint recognition. When the target user identity is a registered user, perform the voice operation corresponding to the target audio in the target user identity. Compared with the related art, the embodiments of the present disclosure determine the corresponding target audio through audio extraction and sound area separation, and determine the target user identity corresponding to the target audio through voiceprint recognition, which can enable the speech recognition application to accurately confirm the user identity of the voice for interaction. In a multi-person speech interaction scenario, accurate speech interaction can be performed.

[0065] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become readily understood from the following description. Description of the Drawings

[0066] The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. Among them:

[0067] Figure 1 is a schematic flowchart of a method for speech recognition provided by an embodiment of the present disclosure;

[0068] Figure 2 is a schematic structural diagram of a separation structure of audio information provided by an embodiment of the present disclosure;

[0069] Figure 3 is a schematic flowchart of a voiceprint registration provided by an embodiment of the present disclosure;

[0070] Figure 4 is a schematic structural diagram of a device for speech recognition provided by an embodiment of the present disclosure;

[0071] Figure 5 is a schematic structural diagram of another device for speech recognition provided by an embodiment of the present disclosure;

[0072] Figure 6 is a schematic block diagram of an electronic device provided by an embodiment of the present disclosure. Detailed Embodiments

[0073] The following describes exemplary embodiments of the present disclosure with reference to the drawings. Various details of the embodiments of the present disclosure are included to assist in understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted below.

[0074] The following describes the method and device for speech recognition, electronic device, and storage medium of the embodiments of the present disclosure with reference to the drawings.

[0075] Figure 1 is a schematic flowchart of a method for speech recognition provided by an embodiment of the present disclosure.

[0076] As Figure 1 shown, the method includes the following steps:

[0077] Step 101: Obtain the speech to be recognized collected by the target speech collection device, perform audio extraction processing on the speech to be recognized, and separate multiple audio corresponding to different users from the speech to be recognized. The speech to be recognized contains audio corresponding to different users, and the audio corresponding to different users is different.

[0078] The present disclosure can be applied to any type of multi-person speech interaction scenarios such as vehicles, homes, work, etc., and any type of multi-person speech interaction games such as shooting, sports, etc. Specifically, the types of application scenarios and games are not limited in the embodiments of the present disclosure.

[0079] In the embodiments of the present disclosure, the speech to be recognized is a kind of speech sentence, the audio is the sentence information containing speech instructions, and the speech control instruction is an instruction to control a device or perform a specific task by a voice command. For example: in a game, instructions such as jumping up, running forward, running backward, etc., which can control the game; in a meeting, instructions such as recording the meeting content to control the speech recognition application, etc. Specifically, the content of the speech to be recognized and the speech control instruction is not limited in the embodiments of the present disclosure.

[0080] Among them, the target speech collection device is a device set by oneself, such as: wireless microphone, wired microphone, sound collecting hole, etc.; performing audio extraction on the speech to be recognized refers to the process of separating different audio from the speech to be recognized containing audio corresponding to different users. For example: separating different audio through different sound frequency bands, etc. Specifically, the target speech collection device and the method of audio extraction are not limited in the embodiments of the present disclosure.

[0081] Step 102: Select the audio information in the target sound area corresponding to the target speech collection device from the multiple audio as the target audio. The target sound area is a pre-divided physical space corresponding to the target speech collection device.

[0082] In the embodiments of the present disclosure, it is necessary to separate the audio information through the sound area of the target speech collection device to obtain the target audio corresponding to the target speech collection device. Regarding the sound area separation of the target speech collection device, it is mainly carried out through the sound parameters of the audio information. For example: the sound intensity, amplitude, etc. of the audio information. The target sound area is a self-defined physical space. For example: defining a circle with a radius of 1 meter around the target speech collection device as the target sound area, etc.

[0083] To facilitate understanding of the implementation process of the embodiments of the present disclosure, an example is provided for illustration: As Figure 2 shown, Figure 2Schematic diagram of a separation structure for audio information provided by an embodiment of the present disclosure. In a vehicle interior space, there are four seats (A, B, C, D), and there is a microphone beside each of the four seats. The audio regions of the four microphones are pre-divided, and each seat corresponds to the audio region of the microphone beside it: There is one person (Jia, Yi, Bing, Ding) in each of the seats (A, B, C, D) now. If the microphone of A is used as the target voice collection device, when Jia and Yi speak simultaneously, the voices of Jia and Yi will be recorded by all four microphones. Separating the target audio from the multiple audio refers to separating two different audio from the voice recorded by the microphone of A. Since Jia is in the audio region of the microphone of A, the audio corresponding to Jia is the target audio, and the audio corresponding to Yi is not the target audio and will not be separated; if the microphone of B is used as the target voice collection device, separating two different audio from the voice recorded by B. Since Yi is in the audio region of the microphone of B, the audio corresponding to Yi is the target audio, and the audio corresponding to Jia is not the target audio and will not be separated; if the microphone of C or the microphone of D is used as the target voice collection device, since neither Jia nor Yi is in the audio region of the microphone of C or the microphone of D, the audio corresponding to Jia and Yi are not target audio and will not be separated. At the same time, the voices recorded by each will be marked as coming from which microphone.

[0084] Step 103: Perform voiceprint recognition on the target audio, and determine the target user identity corresponding to the target audio according to the result of the voiceprint recognition. In the case where the target user identity is a registered user, perform the voice operation corresponding to the target audio in the identity of the target user.

[0085] In an embodiment of the present disclosure, when performing voiceprint recognition, it is performed through a preset voiceprint extraction algorithm. The preset voiceprint extraction algorithm is an algorithm selected customarily, for example, various algorithms corresponding to SV (Speaker Verification, voiceprint recognition), etc. Specifically, regarding the selection of the preset voiceprint extraction algorithm, the embodiment of the present disclosure does not limit it.

[0086] Performing the voice operation corresponding to the target audio in the identity of the target user means performing the voice operation corresponding to the target audio in the identity of a registered user. In addition to performing the voice operation corresponding to the target audio, system operations corresponding to the identity of the registered user will also be performed. Among them, due to the difference between registered and unregistered voiceprints, different system operation rules also correspond. Regarding the content of performing system operations corresponding to the identity of the registered user, for example: sending a welcome message to the user corresponding to the registered voiceprint, blocking users with unregistered voiceprints from performing voice operations, agreeing to users with unregistered voiceprints to perform voice operations, etc. Specifically, regarding the content of the preset voiceprint processing rules, it can be customarily edited according to the actual situation, and the embodiment of the present disclosure does not limit it.

[0087] It should be noted here that the system operations corresponding to the identities of the registered users are additional operations on the system. For example, operations on in-vehicle infotainment systems, game consoles, computers, etc., which have a higher priority than the voice operations corresponding to the target audio and can turn the system on or off. The preset voiceprint database is a custom-built database for storing voiceprint information. Only when the voiceprint recognition result finds a corresponding voiceprint in the preset voiceprint database can it be determined that the target user corresponding to the target audio is a registered user.

[0088] The method for voice recognition provided by the present disclosure includes: obtaining the to-be-recognized voice collected by a target voice collection device, performing audio extraction processing on the to-be-recognized voice, separating multiple audio corresponding to different users from the to-be-recognized voice, where the to-be-recognized voice contains audio corresponding to different users and the audio corresponding to different users is different; selecting, from the multiple audio, the audio in a target sound area corresponding to the target voice collection device as the target audio, where the target sound area is a physically divided space corresponding to the target voice collection device; performing voiceprint recognition on the target audio, determining the identity of the target user corresponding to the target audio according to the result of the voiceprint recognition, and when the identity of the target user is a registered user, performing the voice operation corresponding to the target audio in the identity of the target user. Compared with the related art, the embodiments of the present disclosure determine the corresponding target audio through audio extraction and sound area separation, and determine the identity of the target user corresponding to the target audio through voiceprint recognition, which can enable the voice recognition application to accurately confirm the identity of the user of the voice for interaction and can accurately perform voice interaction in a multi-person voice interaction scenario.

[0089] In an implementable manner of the embodiments of the present disclosure, the multiple audio are extracted from all the sounds that can be received by the target voice collection device, i.e., the to-be-recognized voice. The to-be-recognized voice contains the voice information of multiple users, etc. Through the acquisition of the target audio, voice recognition and multiple sound areas can be combined to improve the effect of voice recognition. Therefore, in order to improve the effect of voice recognition, the following methods can also be used but are not limited to: determining the sound parameters corresponding to each audio information in the multiple audio, where the sound parameters are at least used to characterize the volume of the corresponding audio; based on the sound parameters and the position information of the target voice collection device, locating the sound source of each audio and determining the spatial position of the sound source corresponding to each audio information; and determining the audio with the spatial position of the sound source in the target sound area as the target audio.

[0090] In the embodiments of the present disclosure, the sound parameter is an index used to describe and quantify the characteristic of whether the sound is within the sound area of the target voice collection device. For example, the volume of the audio, the sound intensity of the audio, and the sound amplitude of the audio. When the magnitude of the sound parameter and the position information of the target voice collection device are known, the spatial position of the sound source corresponding to the audio information can be determined. For example, if the target voice collection device is installed in the driver's seat, the sound parameter of the audio emitted by the driver in the driver's seat is relatively large, and the sound parameters of the audio information emitted by people in other positions are relatively small. By comparing the magnitude of the sound parameter with the range of sound parameters received by the target voice collection device at each position, the spatial position of the sound source corresponding to the audio information can be determined. When the spatial position of the sound source corresponding to the audio is within the sound area of the target voice collection device, the audio information can be determined as the target audio.

[0091] In an implementable manner of the embodiments of the present disclosure, related to the above embodiments, since there may be an overlapping area between the sound areas of the target voice collection device and other voice collection devices, when determining the sound area where the spatial position of the sound source is located, it is necessary to accurately identify the sound area where the spatial position of the sound source is located through the distances between the spatial position of the sound source and the target voice collection device and other voice collection devices. Therefore, in order to ensure the accuracy of identifying the sound area where the spatial position of the sound source is located, the following methods can be used but are not limited to: If there is an overlapping area between the target sound area and other sound areas of other voice collection devices, when it is detected that the spatial position of the sound source is in the overlapping area, calculate the distances between the sound source and the target voice collection device and other voice collection devices respectively; when the distance between the sound source in the overlapping area and the target voice collection device is the shortest, determine the audio corresponding to the sound source as the target audio.

[0092] In an implementable manner of the embodiments of the present disclosure, it is necessary to determine whether the voice control instruction in the target audio is valid. Determine whether the voice control instruction corresponding to the target audio is valid. The valid instruction needs to be further executed, and the invalid instruction is directly deleted to reduce the pressure on the voice recognition system. Therefore, in order to save resource waste, the following methods can be used but are not limited to: perform semantic recognition on the target audio to obtain the semantic recognition result corresponding to the target audio; match the semantic recognition result in a preset instruction library; if a corresponding voice control instruction is matched in the preset instruction library, extract the voiceprint feature of the target audio to obtain the corresponding target voiceprint information, and perform voiceprint recognition on the target audio according to the target voiceprint information; if no corresponding voice control instruction is matched in the preset instruction library, determine the target audio as an invalid audio and delete the invalid audio.

[0093] In the embodiments of the present disclosure, the method for semantic recognition of the target audio is an algorithm selected customarily. For example, it can be an algorithm in Automatic Speech Recognition (ASR). Specifically, the embodiments of the present disclosure do not limit the method for semantic recognition of the target audio.

[0094] When performing semantic recognition, it can be carried out by, but not limited to, keyword recognition. For example, during a game, the game name is used as a keyword. After the game name is spoken, the subsequent voice information is the voice control command to be recognized. Through the voice control command, corresponding voice operations are performed. The preset instruction library is the instruction library built into the voice recognition application that needs to be operated. For example, in a chess game, instructions for controlling various chess pieces, etc. Among them, regarding matching corresponding voice control instructions, for example, in a chess game, if the semantic recognition result is that the cannon moves forward two steps, since the cannon moving forward two steps is an instruction existing in the preset instruction library, a corresponding voice control instruction can be matched; if the semantic recognition result is that the pawn moves forward two steps, since the pawn moving forward two steps is not an instruction existing in the preset instruction library, a corresponding voice control instruction cannot be matched. Specifically, the content of the preset instruction library depends on the actual situation, and the embodiments of the present disclosure do not limit it.

[0095] In an implementable manner of the embodiments of the present disclosure, due to the different system operation rules corresponding to registered and unregistered users of voiceprints, it is necessary to identify the voiceprint and determine the target user identity corresponding to the target audio according to the voiceprint recognition result. To distinguish between registered and unregistered users, it can be achieved by, but not limited to, the following methods: matching the target voiceprint information from a preset voiceprint database; if the voiceprint corresponding to the target voiceprint information is matched, it is determined that the target user identity corresponding to the target audio is a registered user; if the voiceprint corresponding to the target voiceprint information is not matched, it is determined that the target user identity corresponding to the target audio is an unregistered user.

[0096] In the embodiments of the present disclosure, after determining that the target user identity corresponding to the target audio is an unregistered user, the system operations corresponding to the identity of the unregistered user will be executed. Among them, the content of the system operations corresponding to the identity of the unregistered user, for example, a prompt of inoperability through a playback device or a display device, a prompt for registering a voiceprint through a playback device or a display device, no response, etc. Specifically, the embodiments of the present disclosure do not limit the content of the system operations corresponding to the identity of the unregistered user.

[0097] It should be noted here that the system operations corresponding to the identity of the unregistered user are also additional operations on the system, such as a car machine, a game console, a computer, etc.

[0098] In an implementable manner of the embodiments of the present disclosure, since the speech to be recognized contains various sounds, that is, the audio corresponding to different users contained in the speech to be recognized, and also contains other speech in addition to the audio information, such as: some other speech unrelated to the voice control instruction, ordinary conversation sounds, discussion sounds with friends, etc.; therefore, it is necessary to perform voice activation processing on the speech to be recognized, that is, the processing of removing noise and detecting human voices, to ensure that there is audio that can be processed in the speech to be recognized. To ensure that audio that can process the speech to be recognized can be obtained, the following methods can also be used but are not limited to: If there is a sound within the preset human voice spectrum range, and it is determined that there is also a sound in the speech to be recognized that is not within the preset human voice spectrum range, then the sound that is not within the preset human voice spectrum range is determined as noise, and the noise in the speech to be recognized is eliminated, and audio extraction processing is performed on the speech to be recognized after eliminating the noise; If there is a sound within the preset human voice spectrum range, and it is determined that there is no sound in the speech to be recognized that is not within the preset human voice spectrum range, then the audio extraction processing is directly performed on the speech to be recognized; If there is no sound within the preset human voice spectrum range, then the sound in the speech to be recognized is determined as noise, and the speech to be recognized is deleted.

[0099] In the embodiments of the present disclosure, the method for detecting the speech to be recognized is a speech detection algorithm selected by customization, such as: various algorithms corresponding to VAD (Voice Activity Detection), etc. Specifically, the present disclosure embodiments do not limit the selection of the preset speech detection algorithm.

[0100] It should be noted here that when human voices are detected in the speech to be recognized, noise elimination processing is required. When it is detected that there are no human voices in the speech to be recognized, the speech to be recognized is invalid speech, and no other operations need to be performed, and it is directly deleted. Among them, the human voice spectrum range contains important information during speaking, such as syllables, intonations, etc., and is the main frequency band analyzed in speech recognition, such as: between 20 Hz and 20,000 Hz. The noise spectrum range refers to the frequency range of any unwanted sound that interferes with normal auditory perception, and can be very wide, and may include frequencies from dozens of hertz to tens of thousands of hertz. In ambient noise, the spectral distribution of noise from different sources will be different. For example: traffic noise may be mainly distributed in the low-frequency band, while the noise of electronic devices may be in the higher frequency band. Specifically, the present disclosure embodiments do not limit the human voice spectrum range and the noise spectrum range.

[0101] In an implementable manner of the embodiments of the present disclosure, the preset voiceprint database is an important database. Only when a corresponding voiceprint is found in the preset voiceprint database for the voiceprint recognition result can it be determined that the target user identity corresponding to the target audio is a registered user, and higher-priority voice operations and system operations can be performed. Therefore, in order to construct the preset voiceprint database, the following methods can be adopted but are not limited to: in response to a voiceprint registration instruction, receive the corresponding registration voiceprint; perform voiceprint registration according to the voiceprint feature vector in the registration voiceprint, and determine whether the voiceprint registration meets the preset voiceprint registration requirements; when it is determined that the number of times the voiceprint registration meets the preset voiceprint registration requirements is greater than or equal to the first preset threshold, store the registration voiceprint in the preset voiceprint database to complete the construction of the preset voiceprint database.

[0102] In the embodiments of the present disclosure, the voiceprint feature vector is a series of numerical features extracted from human speech in voiceprint recognition technology, which can represent the characteristics of an individual's voice. The voiceprint feature vector can reflect the unique attributes of the speaker's voice, such as pitch, timbre, speech rate, accent, etc. The preset voiceprint registration requirements are a kind of requirements set by customization. For example, whether the registration is successful, whether the registered voiceprint can be recognized, etc. Specifically, the embodiments of the present disclosure do not limit this.

[0103] The specific registration process can also be expressed as: in response to a voiceprint registration instruction, control to start the voiceprint registration interface; play a preset broadcast message through the voice playback device in the voiceprint registration interface, and receive the registration voiceprint corresponding to the preset broadcast message; perform voiceprint registration in the voiceprint registration interface according to the registration voiceprint, and determine whether the voiceprint registration is successful; when it is determined that the number of times the voiceprint registration is successful is greater than or equal to the first preset threshold, open the user editing interface; in response to the registration completion instruction triggered by the user in the user editing interface, store the registration voiceprint in the preset voiceprint database to complete the construction of the preset voiceprint database.

[0104] In the embodiments of the present disclosure, the first preset threshold is a value set by customization. For example: 3, 4, 5, etc. Specifically, the embodiments of the present disclosure do not limit the first preset threshold.

[0105] It should be noted here that the initiation of the voiceprint registration instruction and the registration completion instruction can be controlled by voice or manually. Regarding the initiation of the voiceprint registration instruction, for example, when controlled by voice, one can directly say: Ideal Student, make friends, and the voiceprint registration instruction will be initiated; when controlled manually, one can click the make friends option in the settings to initiate the voiceprint registration instruction. Regarding the initiation of the registration completion instruction, for example, when controlled by voice, one can directly say: Complete registration, and the registration completion instruction will be initiated; when controlled manually, one can click the complete option in the user editing interface to initiate the registration completion instruction.

[0106] It should be noted that in the user editing interface, the user's identity information needs to be edited to bind it with the voiceprint information to complete the voiceprint registration. Regarding the content editing in the user editing interface, it can be manually input or voice input. Specifically, the present disclosure embodiment does not limit the way of content editing in the user editing interface.

[0107] Related to the above embodiment, after determining whether the voiceprint registration is successful, the following methods can be used but are not limited to: in the case of determining that the voiceprint registration fails, perform a re-registration operation, play the preset broadcast information again through the voice playback device in the voiceprint registration interface, and receive the registered voiceprint corresponding to the preset broadcast information; if the number of re-registration operations is greater than or equal to the second preset threshold, determine that the voiceprint registration fails and end the voiceprint registration process; in the case of determining that the voiceprint registration is successful, reset the number of re-registration operations.

[0108] In the present disclosure embodiment, the second preset threshold is a value set by the user, for example: 3, 4, 5, etc. Specifically, the present disclosure embodiment does not limit the second preset threshold.

[0109] At the same time, to facilitate the understanding of the above embodiment and the implementation process of the present disclosure embodiment, the present disclosure embodiment provides a flow diagram of voiceprint registration, as Figure 3 shown, in which both the first preset threshold and the second preset threshold are set to 3.

[0110] In one implementable manner of the embodiments of the present disclosure, due to the difference between registered and unregistered voiceprints, different system operations are also corresponding. Among them, the manipulation priority corresponding to registered users is higher than that corresponding to unregistered users. Therefore, in order to ensure the priority of the manipulation corresponding to registered users, the following methods can be used but are not limited to: If the target audio contains at least one target audio corresponding to a registered user and a target audio corresponding to an unregistered user, the target audio corresponding to the registered user is preferentially executed.

[0111] In summary, the embodiments of the present disclosure can achieve the following effects:

[0112] Through audio extraction and voice zone separation, the embodiments of the present disclosure determine the target user identity corresponding to the target audio, which can enable the voice recognition application to accurately confirm the user identity of the voice for interaction. In a multi-person voice interaction scenario, voice interaction can be accurately performed.

[0113] Corresponding to the above method of voice recognition, the present invention also proposes a device for voice recognition. Since the device embodiments of the present invention correspond to the above method embodiments, for the details not disclosed in the device embodiments, reference can be made to the above method embodiments, and details will not be repeated in the present invention.

[0114] Figure 4 As shown in the structural schematic diagram of a device for voice recognition provided by the embodiments of the present disclosure, Figure 4 it includes:

[0115] An extraction unit 41, configured to obtain the speech to be recognized collected by a target speech collection device, perform audio extraction processing on the speech to be recognized, and separate multiple audio corresponding to different users from the speech to be recognized. The speech to be recognized contains audio corresponding to different users, and the audio corresponding to different users is different;

[0116] A selection unit 42, configured to select the audio information in the target voice zone corresponding to the target speech collection device from the multiple audio as the target audio. The target voice zone is a pre-divided physical space corresponding to the target speech collection device;

[0117] An identification unit 43, configured to perform voiceprint recognition on the target audio, and determine the target user identity corresponding to the target audio according to the result of the voiceprint recognition;

[0118] An operation unit 44, configured to execute the voice operation corresponding to the target audio in the identity of the target user when the target user identity is a registered user.

[0119] The voice recognition device provided by the present disclosure obtains the voice to be recognized collected by the target voice collection device, performs audio extraction processing on the voice to be recognized, separates multiple audio corresponding to different users from the voice to be recognized, where the voice to be recognized contains audio corresponding to different users, and the audio corresponding to different users is different; selects the audio in the target sound area corresponding to the target voice collection device from the multiple audio as the target audio, where the target sound area is a pre-divided physical space corresponding to the target voice collection device; performs voiceprint recognition on the target audio, determines the target user identity corresponding to the target audio according to the result of the voiceprint recognition, and when the target user identity is a registered user, executes the voice operation corresponding to the target audio in the target user identity. Compared with the related art, the embodiments of the present disclosure determine the corresponding target audio through audio extraction and sound area separation, and determine the target user identity corresponding to the target audio through voiceprint recognition, which can enable the voice recognition application to accurately confirm the user identity of the voice for interaction, and can accurately perform voice interaction in a multi-person voice interaction scenario.

[0120] Further, in a possible implementation manner of the embodiments of the present disclosure, the selection unit 42 is further configured to:

[0121] Determine the sound parameters corresponding to each audio in the multiple audio, where the sound parameters are at least used to characterize the volume size of the corresponding audio;

[0122] Based on the sound parameters and the position information of the target voice collection device, locate the sound source of each audio, and determine the spatial position of the sound source corresponding to each audio;

[0123] Determine the audio with the spatial position of the sound source in the target sound area as the target audio.

[0124] Further, in a possible implementation manner of the embodiments of the present disclosure, the selection unit 42 is further configured to:

[0125] When there is an overlapping area between the target sound area and other sound areas of other voice collection devices, in the case of detecting that the spatial position of the sound source is in the overlapping area, calculate the distances between the sound source and the target voice collection device and other voice collection devices respectively;

[0126] When the distance between the sound source in the overlapping area and the target voice collection device is the shortest, determine the audio corresponding to the sound source as the target audio.

[0127] Further, in a possible implementation manner of the embodiments of the present disclosure, as Figure 5 shown, the device further includes:

[0128] The recognition unit 43 is further configured to perform semantic recognition on the target audio to obtain a semantic recognition result corresponding to the target audio;

[0129] A matching unit 45 is configured to match the semantic recognition result in a preset instruction library;

[0130] The recognition unit 43 is further configured to, when a corresponding voice control instruction is matched in the preset instruction library, extract a voiceprint feature from the target audio to obtain corresponding target voiceprint information, and perform voiceprint recognition on the target audio according to the target voiceprint information;

[0131] The recognition unit 43 is further configured to, when no corresponding voice control instruction is matched in the preset instruction library, determine that the target audio is an invalid audio and delete the invalid audio.

[0132] Further, in a possible implementation manner of the embodiments of the present disclosure, as Figure 5 shown, the recognition unit 43 includes:

[0133] A matching module 431 is configured to match the target voiceprint information in a preset voiceprint database;

[0134] A determination module 432 is configured to determine that the target user identity corresponding to the target audio is a registered user when the voiceprint corresponding to the target voiceprint information is matched;

[0135] The determination module 432 is further configured to determine that the target user identity corresponding to the target audio is an unregistered user when the voiceprint corresponding to the target voiceprint information is not matched.

[0136] Further, in a possible implementation manner of the embodiments of the present disclosure, as Figure 5 shown, the device further includes:

[0137] A detection unit 46 is configured to detect whether there is a sound within a preset human voice spectrum range in the voice to be recognized, where the preset human voice spectrum range is the spectrum range of the user's voice;

[0138] A clearing unit 47 is configured to, when there is a sound within the preset human voice spectrum range and it is determined that there is also a sound not within the preset human voice spectrum range in the voice to be recognized, determine the sound not within the preset human voice spectrum range as noise, eliminate the noise in the voice to be recognized, and perform audio extraction processing on the voice to be recognized after the noise is eliminated;

[0139] The extraction unit 41 is further configured to directly perform audio extraction processing on the speech to be recognized when there is a sound within a preset human voice spectrum range and it is determined that there is no sound in the speech to be recognized that is not within the preset human voice spectrum range.

[0140] The deletion unit 48 is configured to determine that the sound in the speech to be recognized is noise and delete the speech to be recognized when there is no sound within the preset human voice spectrum range.

[0141] Further, in a possible implementation manner of the embodiments of the present disclosure, the operation unit 44 is further configured to, when the target audio includes at least one target audio corresponding to a registered user and a target audio corresponding to an unregistered user, preferentially execute the voice operation corresponding to the target audio corresponding to the registered user.

[0142] It should be noted that the foregoing explanations of the method embodiments also apply to the devices of the embodiments of the present disclosure. The principles are the same and are not limited again in the embodiments of the present disclosure.

[0143] According to the embodiments of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0144] Figure 6 FIG. shows a schematic block diagram of an exemplary electronic device 600 that can be used to implement the embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as, for example, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, for example, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely exemplary and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0145] As Figure 6 shown, the device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to a computer program stored in a ROM (Read-Only Memory) 602 or a computer program loaded from a storage unit 608 into a RAM (Random Access Memory) 603. In the RAM 603, various programs and data required for the operation of the device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other through a bus 604. An I / O (Input / Output) interface 605 is also connected to the bus 604.

[0146] Multiple components in device 600 are connected to I / O interface 605, including: input unit 606, such as a keyboard, mouse, etc.; output unit 607, such as various types of displays, speakers, etc.; storage unit 608, such as a disk, optical disc, etc.; and communication unit 609, such as a network card, modem, wireless communication transceiver, etc. Communication unit 609 allows device 600 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0147] Computing unit 601 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of computing unit 601 include but are not limited to CPU (Central Processing Unit), GPU (Graphic Processing Units), various dedicated AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSP (Digital Signal Processor), and any suitable processor, controller, microcontroller, etc. Computing unit 601 executes the various methods and processes described above, such as the method of speech recognition. For example, in some embodiments, the method of speech recognition can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed onto device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by computing unit 601, one or more steps of the method described above can be executed. Alternatively, in other embodiments, computing unit 601 can be configured to execute the aforementioned method of speech recognition in any other suitable way (e.g., by means of firmware).

[0148] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuitry, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application Specific Standard Products), SoCs (System On Chip), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that receives data and instructions from a storage system, at least one input device, and at least one output device, and transmits the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0149] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowchart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0150] In the context of this disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a RAM, a ROM, an EPROM (Electrically Programmable Read-Only Memory), or a flash memory, an optical fiber, a CD-ROM (Compact Disc Read-Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0151] To provide for interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide for interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, speech input, or tactile input).

[0152] The systems and techniques described herein can be implemented in a computing system that includes backend components (such as, for example, a data server), or a computing system that includes middleware components (such as, for example, an application server), or a computing system that includes frontend components (such as, for example, a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (such as, for example, a communication network). Examples of communication networks include: a LAN (Local Area Network), a WAN (Wide Area Network), the Internet, and a blockchain network.

[0153] A computer system may include a client and a server. The client and the server are generally far from each other and usually interact via a communication network. The relationship between the client and the server is generated by computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, solving the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services ("Virtual Private Server", or simply "VPS"). The server may also be a server of a distributed system or a server combined with a blockchain.

[0154] Among them, it should be noted that artificial intelligence is a discipline that studies how to make a computer simulate certain human thinking processes and intelligent behaviors (such as learning, reasoning, thinking, planning, etc.), and it has technologies at both the hardware level and the software level. Artificial intelligence hardware technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, and big data processing; artificial intelligence software technologies mainly include several major directions such as computer vision technology, speech recognition technology, natural language processing technology, and machine learning / deep learning, big data processing technology, and knowledge graph technology.

[0155] It should be understood that various forms of processes shown above can be used, and steps can be reordered, added, or deleted. For example, the steps recorded in this disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this is not limited herein.

[0156] The above specific implementation manners do not constitute a limitation on the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the protection scope of this disclosure.

Claims

1. A method for speech recognition, characterized in that, Including: Obtain the speech to be recognized collected by the target speech collection device, perform audio extraction processing on the speech to be recognized, and separate multiple audio corresponding to different users from the speech to be recognized. The speech to be recognized contains audio corresponding to different users, and the audio corresponding to different users is different; From the multiple audio, select the audio in the target sound area corresponding to the target speech collection device as the target audio. The target sound area is a pre-divided physical space corresponding to the target speech collection device; Perform voiceprint recognition on the target audio, determine the target user identity corresponding to the target audio according to the result of the voiceprint recognition, and execute the voice operation corresponding to the target audio in the case that the target user identity is a registered user.

2. The method according to claim 1, wherein The step of selecting the audio in the target sound area corresponding to the target speech collection device from the multiple audio includes: Determine the sound parameters corresponding to each audio in the multiple audio, and the sound parameters are at least used to characterize the volume size of the corresponding audio; Based on the sound parameters and the position information of the target speech collection device, locate the sound source of each audio and determine the spatial position of the sound source corresponding to each audio; Determine the audio with the spatial position of the sound source in the target sound area as the target audio.

3. The method according to claim 2, characterized in that The step of determining the audio with the spatial position of the sound source in the target sound area as the target audio includes: If there is an overlapping area between the target sound area and other sound areas of other speech collection devices, then in the case of detecting that the spatial position of the sound source is in the overlapping area, calculate the distances between the sound source and the target speech collection device and other speech collection devices respectively; In the case that the distance between the sound source in the overlapping area and the target speech collection device is the shortest, determine the audio corresponding to the sound source as the target audio.

4. The method according to claim 1, wherein Before performing voiceprint recognition on the target audio, it further includes: Perform semantic recognition on the target audio to obtain the semantic recognition result corresponding to the target audio; Match the semantic recognition result in a preset instruction library; If a corresponding voice control instruction is matched in the preset instruction library, then perform voiceprint feature extraction on the target audio to obtain the corresponding target voiceprint information, and perform voiceprint recognition on the target audio according to the target voiceprint information; If no corresponding voice control instruction is matched in the preset instruction library, then determine the target audio as invalid audio and delete the invalid audio.

5. The method according to claim 4, characterized in that, The step of determining the target user identity corresponding to the target audio according to the result of the voiceprint recognition includes: Match the target voiceprint information in a preset voiceprint database; If the voiceprint corresponding to the target voiceprint information is matched, then determine that the target user identity corresponding to the target audio is a registered user; If the voiceprint corresponding to the target voiceprint information is not matched, then determine that the target user identity corresponding to the target audio is an unregistered user.

6. The method according to claim 1, wherein Before performing audio extraction processing on the speech to be recognized, it further includes: Detect whether there is a sound within a preset human voice spectrum range in the to-be-recognized speech, where the preset human voice spectrum range is the spectrum range of the user's voice; If there is a sound within the preset human voice spectrum range, and if it is determined that there is also a sound in the to-be-recognized speech that is not within the preset human voice spectrum range, then determine the sound that is not within the preset human voice spectrum range as noise, eliminate the noise in the to-be-recognized speech, and perform audio extraction processing on the to-be-recognized speech after noise elimination; If there is a sound within the preset human voice spectrum range, and if it is determined that there is no sound in the to-be-recognized speech that is not within the preset human voice spectrum range, then directly perform audio extraction processing on the to-be-recognized speech; If there is no sound within the preset human voice spectrum range, then determine the sound in the to-be-recognized speech as noise and delete the to-be-recognized speech.

7. The method according to claim 1, characterized in that, The method further includes: When the target audio includes the audio corresponding to a registered user and the audio corresponding to an unregistered user, preferentially execute the voice operation corresponding to the audio corresponding to the registered user.

8. A device for speech recognition, characterized in that, It includes: An extraction unit, configured to obtain the to-be-recognized speech collected by a target voice collection device, perform audio extraction processing on the to-be-recognized speech, and separate multiple audios corresponding to different users from the to-be-recognized speech. The to-be-recognized speech contains audios corresponding to different users, and the audios corresponding to different users are different; A selection unit, configured to select, from the multiple audios, the audio within the target sound area corresponding to the target voice collection device as the target audio, where the target sound area is a pre-divided physical space corresponding to the target voice collection device; An identification unit, configured to perform voiceprint identification on the target audio, and determine the target user identity corresponding to the target audio according to the result of the voiceprint identification; An operation unit, configured to, when the target user identity is a registered user, execute the voice operation corresponding to the target audio in the identity of the target user.

9. An electronic device, characterized in that, It includes: At least one processor; And A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1-7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to execute the method according to any one of claims 1-7.

Images