Identity identification method, device and equipment and storage medium

Abstract

Embodiments of the present application disclose an identity recognition method, device and equipment, and a storage medium, and belong to the technical field of computers. The method comprises: obtaining a voiceprint recognition result of a target object in a conference based on a voice of the target object, the voiceprint recognition result being used to indicate whether the voiceprint of the target object has been registered; in the case where the voiceprint recognition result indicates that the voiceprint of the target object has not been registered, obtaining conference information of the conference; and performing identity recognition on the target object based on the conference information. The above technical solution can realize automatic recognition of the identity of a user without the user actively inputting a voiceprint, thereby greatly improving the identity recognition efficiency.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to an identity recognition method, apparatus, device, and storage medium. Background Technology

[0002] With the rapid development of computer technology, speaker identification technology has been widely applied in various aspects of daily life. Currently, in meeting scenarios, speaker identification is mainly based on the speaker's voiceprint. However, in order to match a speaker's voiceprint with the target identity, users often need to actively register their voiceprints before the meeting and store them in a voiceprint database. When there are many participants, this process becomes quite cumbersome, resulting in low efficiency in speaker identification. Summary of the Invention

[0003] This application provides an identity recognition method, apparatus, device, and storage medium that can automatically recognize user identity without requiring the user to actively record their voiceprint, thereby greatly improving identity recognition efficiency. The technical solution is as follows:

[0004] On the one hand, an identity recognition method is provided, the method comprising:

[0005] Based on the speech of the target object in the meeting, the voiceprint recognition result of the target object is obtained, and the voiceprint recognition result is used to indicate whether the voiceprint of the target object has been registered;

[0006] If the voiceprint recognition result indicates that the target object's voiceprint is not registered, obtain the meeting information of the meeting;

[0007] Based on the meeting information, the target object is identified.

[0008] On the other hand, an identity recognition device is provided, the device comprising:

[0009] The first acquisition module is used to acquire the voiceprint recognition result of the target object based on the voice of the target object in the meeting, and the voiceprint recognition result is used to indicate whether the voiceprint of the target object has been registered.

[0010] The second acquisition module is used to acquire meeting information of the meeting when the voiceprint recognition result indicates that the voiceprint of the target object is not registered;

[0011] The identity recognition module is used to identify the target object based on the meeting information.

[0012] In some embodiments, the meeting information includes the voices of other participants in the meeting besides the target object, and the identity recognition module includes:

[0013] The recognition unit is used to perform speech recognition on the speech of the other participants to obtain the speech information of the other participants besides the target object;

[0014] The determining unit is used to determine the speech information related to the target object from the speech information of other participants besides the target object;

[0015] The determining unit is further configured to determine the identity information of the target object based on the speech information related to the target object.

[0016] In some embodiments, the determining unit is configured to: determine, from the speaking information of other participants besides the target object, the speaking information within a first preset time period before the target object speaks; determine, from the speaking information within the first preset time period before the target object speaks, the identity information of the next participant to speak; and determine, as the identity information of the next participant to speak, the identity information of the target object.

[0017] In some embodiments, the determining unit is configured to: determine, from the speaking information of other participants besides the target object, speaking information related to the speaking order of the meeting; determine, from the speaking information related to the speaking order of the meeting, the identity information of the participant corresponding to the speaking order of the target object; and determine the identity information of the participant as the identity information of the target object.

[0018] In some embodiments, the determining unit is configured to determine, from the speaking information of other participants besides the target object, the speaking information within a second preset time period after the target object speaks; determine the identity information of the previous speaker from the speaking information within the second preset time period after the target object speaks; and determine the identity information of the previous speaker as the identity information of the target object.

[0019] In some embodiments, the meeting information includes a first image displayed on a screen during the meeting, the first image including the meeting's display information, and the identity recognition module is used to perform text recognition on the first image to obtain the meeting's display information; determine the identity information of the presenter of the display information from the display information; and determine the identity information of the target object based on the identity information of the presenter of the display information.

[0020] In some embodiments, the identity recognition module is configured to: determine the identity information of the presenter as the identity information of the target object when the presenter of the information is a single person and the speech information of the target object is related to the meeting information; and determine the presenter corresponding to the speech information or speech order of the target object based on the presentation order of the multiple presenters and the presentation information corresponding to each presenter when the presenter of the information is a multiple person, and determine the identity information of the presenter corresponding to the speech information or speech order of the target object as the identity information of the target object.

[0021] In some embodiments, the meeting information includes a second image captured by a camera during the meeting, the second image including all participants in the meeting, and the identity recognition module is used to perform object detection on the second image to determine a target region, the target region being the region where the target object is located in the second image; based on the target region, at least one of face recognition and text recognition is performed on the second image to obtain the identity information of the target object.

[0022] In some embodiments, the identity recognition module is configured to perform face recognition on the second image based on the target region to obtain a face recognition result of the target object, wherein the face recognition result is used to indicate whether the face of the target object has been registered; if the face recognition result indicates that the face of the target object has been registered, the identity information corresponding to the face is determined as the identity information of the target object.

[0023] In some embodiments, the identity recognition module is configured to perform text recognition on the second image based on the target region to obtain a text recognition result of the target object, wherein the text recognition result is used to indicate whether the target region includes text information for representing identity; and if the text recognition result indicates that the target region includes text information for representing identity, the text information is determined as the identity information of the target object.

[0024] In some embodiments, the identity recognition module is configured to perform face recognition and text recognition on the second image based on the target region to obtain face recognition results and text recognition results of the target object. The face recognition results are used to indicate whether the face of the target object has been registered, and the text recognition results are used to indicate whether the target region includes text information used to represent identity. Based on the face recognition results and the text recognition results, the identity information of the target object is determined.

[0025] In some embodiments, the apparatus further includes:

[0026] The input module is used to input the voiceprint of the target object and the identity information corresponding to the voiceprint into the voiceprint database.

[0027] On the other hand, a computer device is provided, the computer device including a processor and a memory, the memory storing at least one computer program, the at least one computer program being loaded and executed by the processor to implement the identity recognition method as described above.

[0028] On the other hand, a computer-readable storage medium is provided that stores at least one computer program, which is loaded and executed by a processor to implement the identity recognition method as described above.

[0029] On the other hand, a computer program product is provided, including a computer program that is loaded and executed by a processor to implement the identity recognition method as described above.

[0030] This application provides an identity recognition method that can identify the target object in a meeting based on the meeting information when the identity of the target object cannot be identified by voiceprint. This eliminates the need for the user to actively record their voiceprint, thus achieving automatic identification of the user's identity and greatly improving the efficiency of identity recognition. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a schematic diagram of an implementation environment provided in an embodiment of this application;

[0033] Figure 2 This is a flowchart of an identity recognition method provided in an embodiment of this application;

[0034] Figure 3 This is a flowchart of another identity recognition method provided in the embodiments of this application;

[0035] Figure 4 This is an overall flowchart of an identity recognition process provided in an embodiment of this application;

[0036] Figure 5 This is a schematic diagram of a voice content analysis module provided in an embodiment of this application;

[0037] Figure 6This is a schematic diagram of the structure of an identity recognition device provided in an embodiment of this application;

[0038] Figure 7 This is a schematic diagram of another identity recognition device provided in an embodiment of this application;

[0039] Figure 8 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the implementation methods of this application will be further described in detail below with reference to the accompanying drawings.

[0041] It is understood that the terms "first," "second," etc., used in this application may be used herein to describe various concepts, but unless otherwise stated, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of this application, a first image may be referred to as a second image, and similarly, a second image may be referred to as a first image.

[0042] It should be noted that all information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.), and signals involved in this application have been authorized by the user or fully authorized by all parties, and the collection, use, and processing of related data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. For example, the voice and images involved in this application were obtained with full authorization.

[0043] The implementation environment of the embodiments of this application is described below.

[0044] Figure 1 This is a schematic diagram of an implementation environment provided in an embodiment of this application. See also... Figure 1 The implementation environment includes: computer equipment 101, server 102, screen 103 and camera 104.

[0045] Optionally, the computer device 101 can be an all-in-one computer, desktop computer, laptop, smartphone, tablet computer, or other computer device. The computer device 101 and the server 102 are connected via a communication network. Optionally, the computer device 101 is equipped with an intelligent meeting recording application for accessing the server 102, which can automatically identify the participants in the meeting and record what they say.

[0046] Optionally, the computer device 101 is connected to both the screen 103 and the camera 104. The computer device 101 can acquire images displayed on the screen 103 and images captured by the camera 104. The screen 103 is the screen of an information display device used in the meeting, which can be an all-in-one computer, desktop computer, laptop, electronic whiteboard, etc.

[0047] Optionally, server 102 can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0048] Optionally, the identity recognition method provided in this application embodiment can be executed by computer device 101. When computer device 101 executes the identity recognition method, it can run an intelligent meeting recording application offline. When identifying the identity of participants in a meeting, the intelligent meeting recording application can acquire the participants' voices through a built-in or external audio acquisition device of computer device 101, and perform voiceprint recognition on the voices to obtain the participants' voiceprint recognition results. If the voiceprint recognition result indicates that the voiceprint represented by the participant's voice is not registered, then the participants can be identified based on the meeting information, wherein the meeting information includes at least one of the following: the voices of other participants acquired by the audio acquisition device, the images displayed on screen 103, and the images acquired by camera 104.

[0049] Optionally, the identity recognition method provided in this application embodiment can also be interactively executed by computer device 101 and server 102. When computer device 101 and server 102 interactively execute the identity recognition method, the intelligent meeting recording application running on computer device 101 is associated with server 102, and server 102 provides background services. During the operation of the intelligent meeting recording application on computer device 101, the application can acquire the voice of participants through a built-in or external audio acquisition device and send the voice to server 102. Server 102 can perform voiceprint recognition on the voice and return the resulting voiceprint recognition result to computer device 101. If the voiceprint recognition result indicates that the voiceprint represented by the participant's voice is not registered, then the participant's identity can be recognized based on the meeting information. The meeting information includes at least one of the following: voices of other participants acquired by the audio acquisition device, images displayed on screen 103, and images acquired by camera 104.

[0050] Figure 2 This is a flowchart of an identity recognition method provided in an embodiment of this application. This embodiment is executed by a computer device as an example. See also... Figure 2 The method includes:

[0051] 201. The computer device obtains the voiceprint recognition result of the target object based on the voice of the target object in the meeting. The voiceprint recognition result is used to indicate whether the voiceprint of the target object has been registered.

[0052] In this embodiment, the meeting can be a live meeting or a web meeting; this application does not impose any limitation on this. The target object refers to any participant speaking in the meeting, i.e., the speaker in the meeting. During the operation of the computer device, the computer device can acquire the voice of the target object in the meeting through its built-in or external audio acquisition device (such as a microphone). Since the target object's voice contains its voiceprint features, the computer device can obtain the target object's voiceprint recognition result based on the target object's voice. The voiceprint recognition result is used to indicate whether the target object's voiceprint has been registered.

[0053] Optionally, the computer device can perform voiceprint recognition on the target object's speech to obtain the voiceprint recognition result. It can also send the obtained speech of the target object to a server via a network, whereby the server performs voiceprint recognition on the target object's speech and receives the voiceprint recognition result returned by the server. This application embodiment does not limit this.

[0054] Voiceprint recognition involves analyzing the speech of a target individual, extracting voiceprint features, and comparing these features with existing voiceprint templates in a voiceprint database to determine if a matching template exists. If a matching template exists in the database, it indicates that the target individual's voiceprint is registered, and the identity information corresponding to the matching template can be used to identify the target individual. Conversely, if no matching template exists in the database, it indicates that the target individual's voiceprint is not registered.

[0055] 202. If the voiceprint recognition result indicates that the target object's voiceprint is not registered, the computer device obtains the meeting information.

[0056] In this embodiment, when the voiceprint recognition result indicates that the target object's voiceprint is not registered, it means that the target object's identity cannot be identified through its voiceprint. Therefore, the computer device can begin collecting other information to identify the target object, i.e., it can obtain the meeting information. The meeting information includes at least one of the following: voice recordings of other participants besides the target object collected by an audio acquisition device; a first image displayed on the screen obtained using a screenshot function or an image acquisition device connected to the screen in the meeting; and a second image obtained using a camera. The first image includes meeting presentation information, such as the content of a PowerPoint presentation, meeting documents, charts, etc. The second image includes all participants in the meeting, recording their appearance, location, actions, and other information.

[0057] 203. Computer equipment identifies target individuals based on meeting information.

[0058] In this embodiment of the application, since the voices of other participants may contain speaking information related to the target object, or the first image may also include display information related to the target object, or the second image may include not only the facial information of the target object, but also text information related to the target object, the computer device can identify the target object based on meeting information including at least one of the voices of other participants, the first image, and the second image.

[0059] This application provides an identity recognition method that can identify the target object in a meeting based on the meeting information when the identity of the target object cannot be identified by voiceprint. This eliminates the need for the user to actively record their voiceprint, thus achieving automatic identification of the user's identity and greatly improving the efficiency of identity recognition.

[0060] The above Figure 2 The main flow of the identity recognition method provided in the embodiments of this application is illustrated by way of example. The target object identity recognition scheme will be described in detail below. Figure 3 This is a flowchart of another identity recognition method provided in an embodiment of this application. This method is executed by a computer device. See [link to flowchart]. Figure 3 The method includes:

[0061] 301. The computer device obtains the voiceprint recognition result of the target object based on the voice of the target object in the meeting. The voiceprint recognition result is used to indicate whether the voiceprint of the target object has been registered.

[0062] In this embodiment, step 301 is the same as step 201 described above, and will not be repeated here.

[0063] 302. If the voiceprint recognition result indicates that the target object's voiceprint is not registered, the computer device obtains the meeting information.

[0064] In this embodiment, step 302 is the same as step 202 described above, and will not be repeated here.

[0065] It should be noted that the meeting information includes at least one of the following: the voices of other participants besides the target, the first image displayed on the screen, and the second image captured by the camera. Accordingly, when the computer device identifies the target based on the voices of other participants besides the target, please refer to steps 303-304 below; when the computer device identifies the target based on the first image displayed on the screen during the meeting, please refer to steps 305-307 below; when the computer device identifies the target based on the second image captured by the camera during the meeting, please refer to steps 308-309 below.

[0066] 303. The computer equipment performs speech recognition on the speech of participants other than the target, and obtains the speaking information of the participants other than the target.

[0067] In this embodiment, in order to obtain text information related to the current target object from the speech of other participants besides the target object, the computer device needs to perform speech recognition on the speech of other participants to obtain the corresponding text content, thereby obtaining the speech information of other participants. The speech information refers to the content conveyed by the participants through speech, which is presented in text form after speech recognition, including information such as the topics discussed, viewpoints, and people or events mentioned by the participants.

[0068] 304. The computer equipment determines the speech information related to the target from the speech information of other participants besides the target; based on the speech information related to the target, it determines the target's identity information.

[0069] In this embodiment, after acquiring the speaking information of other participants, the computer device can use natural language processing technology to analyze the text content of this speaking information to extract speaking information related to the target object. Speaking information related to the target object refers to parts of the other participants' speaking content that directly or indirectly mention the target object, such as mentioning the target object's name, position, events or behaviors related to the target object, etc. This information helps to determine the target object's identity. Accordingly, the computer device can determine the target object's identity information based on the speaking information related to the target object. Identity information refers to information that clearly identifies the target object's characteristics, such as name, position, department, role in the meeting, etc. This information is used to uniquely identify the target object's identity in a specific environment. By determining speaking information related to the target object from the voices of other participants in the meeting, the target object's identity can be reasonably inferred based on this information, thus eliminating the need for users to actively record voiceprints before the meeting and improving the efficiency of target object identification.

[0070] Optionally, during the analysis, a list of keywords potentially related to the target audience, such as names or job titles, can be set, and matching keywords can be searched in the statements. When a keyword is detected, the overall meaning of the statements related to the keyword can be understood by further analyzing the context of the keyword, thereby determining whether it is relevant to the target audience.

[0071] In some embodiments, step 304 above includes the following three implementation methods:

[0072] Method 1: The computer device determines the speech information within a first preset time period before the target person speaks from the speech information of other participants besides the target person; determines the identity information of the next participant to speak from the speech information within the first preset time period before the target person speaks; and identifies the identity information of the next participant as the identity information of the target person.

[0073] The first preset duration is used to limit the time range for filtering the speaking information of other participants. Its purpose is to obtain relevant information within a specific time period before the target participant speaks, in order to extract clues related to the target participant's identity. The first preset duration can be 5 minutes, 10 minutes, etc. When the computer device obtains the speaking information of other participants, it adds an accurate timestamp to each statement, recording the specific time it occurred. When the target participant begins to speak, it filters all content within the first preset duration from the previous statements based on the recorded time information. For example, if the first preset duration is set to 5 minutes, and the target participant begins speaking after 10 minutes of the meeting, the computer device will extract the speaking information of other participants between 5 and 10 minutes. Then, the computer device uses natural language processing technology to analyze the text content of these statements, understand their meaning, and extract the identity information of the next participant to speak, such as "Next, it's XX's turn," "Next, please let XX speak," "XX, what stage is the project at?" etc. Under normal circumstances, the order of speaking has a certain continuity; that is, the participant mentioned as the next to speak is usually the one who will speak next. Therefore, computer equipment can directly identify the target participant's identity by analyzing the information from the speeches of those who have spoken before them. By first identifying the speeches of those who spoke before the target participant from the speech of other participants, and then using the information from that speech to introduce the next speaker, the identity of the target participant can be reasonably inferred. This allows for identification of the target participant without analyzing all the speeches of other participants, thus further improving the efficiency of target participant identification.

[0074] Method 2: The computer device determines the speaking information related to the speaking order of the meeting from the speaking information of other participants besides the target; from the speaking information related to the speaking order of the meeting, it determines the identity information of the participants corresponding to the speaking order of the target; and determines the identity information of the participants as the identity information of the target.

[0075] The speaking order in a meeting refers to the pre-set or explicitly stated order in which participants speak. When acquiring speaking information from other participants, computer equipment uses natural language processing (NLP) technology to analyze the text content, understand its meaning, and identify specific words and phrases related to the speaking order. These include words indicating sequence such as "first," "then," "next," and "next," as well as phrases explicitly mentioning the speaker's name and the order of content, like "discuss with...". Once sentences matching these characteristics are identified, the computer equipment extracts them from the overall speaking information to obtain the speaking information related to the meeting's speaking order. For example, in a meeting's speaking information, a sentence like "First discuss the first item with Li XX, then discuss the second item with Liu XX" would be filtered out. Then, the computer equipment can match the target's speaking order with the previously analyzed speaking order to determine the identity of the participant corresponding to the target's speaking order. For example, if the currently unknown target is second in the speaking order, then the corresponding participant is Liu XX. After the above analysis, the computer device can determine the identity information of the participants corresponding to the speaking order of the target object as the identity information of the current target object. By first determining the speaking information related to the speaking order from the voices of other participants in the meeting, and then reasonably inferring the identity of the target object based on the participant information corresponding to the speaking order of the target object in that information, the identity of the target object can be determined without analyzing all the speaking information of other participants, thereby further improving the efficiency of target object identification.

[0076] Method 3: The computer device determines the speech information within a second preset time period after the target object speaks from the speech information of other participants besides the target object; determines the identity information of the previous speaker from the speech information within the second preset time period after the target object speaks; and determines the identity information of the previous speaker as the identity information of the target object.

[0077] The second preset duration is used to limit the time range for filtering the speaking information of other participants. Its purpose is to obtain relevant information within a specific time period after the target participant speaks, in order to extract clues related to the target participant's identity. The second preset duration can be 5 minutes, 10 minutes, etc. When the computer device obtains the speaking information of other participants, it adds an accurate timestamp to each message, recording the specific time it occurred. When the moment when the target participant's speech ends is detected, all content generated within the second preset duration is filtered from subsequent messages based on that time. For example, if the second preset duration is set to 5 minutes, and the target participant ends their speech at 15 minutes into the meeting, the computer device will capture the speaking information of other participants between 15 and 20 minutes. Then, the computer device will use natural language processing technology to analyze the text content of these messages, understand their meaning, and extract the identity information of the participant who spoke previously, such as "Thank you for your speech, XX," or "Hello, XX, I have a few questions regarding your speech." Because of the continuity of speeches and typical communication patterns in meetings, after the target participant finishes speaking, the subsequent feedback from other participants is usually directed at the person who just spoke. Therefore, based on this logic, computer equipment can directly identify the identity of the participant who spoke after the target participant by analyzing the information in the subsequent speeches. By first identifying the speeches following the target participant from the audio of other participants, and then using the information in that audio that introduces the previous speaker to reasonably infer the target participant's identity, the target participant's identity can be determined without analyzing all the speeches from other participants, thus further improving the efficiency of target participant identification.

[0078] 305. The computer equipment performs text recognition on the first image to obtain the conference display information.

[0079] In this embodiment, the first image refers to the image displayed on the screen during the meeting. Its content typically includes various materials related to the meeting topic, such as PPT content and document presentations. After acquiring the first image, the computer device can use Optical Character Recognition (OCR) technology to perform text recognition on the first image to obtain the meeting's presentation information. This presentation information is the text content obtained after text recognition of the first image, covering various key information presented during the meeting, such as the meeting topic, discussion content, and presenter information.

[0080] 306. The computer equipment determines the identity information of the presenter from the displayed information.

[0081] In this embodiment, the presenter refers to the person responsible for presenting and explaining the relevant content of the meeting to the participants. Their identity information plays an important role in determining the identity information of the current target. Accordingly, the computer device can use natural language processing technology to analyze the text content of the presented information, understand its meaning, and extract the presenter's identity information, such as "Presenter / Speaker: XX".

[0082] 307. The computer equipment determines the identity information of the target object based on the identity information of the presenter of the displayed information.

[0083] In this embodiment, in most meeting scenarios, the presenter is often one of the main speakers, and their speech is closely related to the information displayed on the first image. Based on this, the computer device can determine that the current target object is the presenter when there is a high correlation between the current target object and the presenter in the displayed information; that is, it can determine the target object's identity information based on the presenter's identity information. By determining the presenter's identity information from the information displayed in the first image, the target object's identity information can be reasonably inferred based on this information, thus eliminating the need for users to actively record their voiceprints before the meeting and improving the efficiency of target object identification.

[0084] In some embodiments, when determining the identity information of the presenter displaying the information, the computer device can also determine the number of presenters. Accordingly, step 307 above includes the following two implementation methods:

[0085] Implementation Method 1: When there is only one presenter and the target audience's speech is relevant to the presented information, the computer device will identify the presenter as the target audience's identity. Specifically, if there is only one presenter and the presented information is relevant to the target audience's speech, based on the logic that the presenter typically leads the discussion in meetings, it can be assumed that the presenter is highly likely to be the target audience speaking on the meeting's theme. Therefore, the computer device can directly identify the presenter's identity as the target audience's identity.

[0086] Method Two: When there are multiple presenters, the computer device determines the presenter corresponding to the target audience's presentation information or presentation order based on the presentation order of the multiple presenters and the presentation information corresponding to each presenter. The identity information of the presenter corresponding to the target audience's presentation information or presentation order is then identified as the target audience's identity information. The presentation order refers to the sequence in which multiple presenters report in the meeting. When there are multiple presenters, the computer device can extract the presentation order of the multiple presenters and the specific presentation information corresponding to each presenter from the presentation information. For example, the presentation information may clearly indicate "Presenter A will first explain Part A; Presenter B will then explain Part B," and the computer device can determine the presentation order and the information section each presenter is responsible for based on this. Next, on the one hand, the computer device can perform a detailed analysis of the target audience's presentation information, extract key information, and then accurately match it with the presentation information corresponding to each presenter to find the presenter whose presentation information best matches the target audience's presentation information. This presenter is the one corresponding to the target audience's presentation information. On the other hand, by matching the speaking order of the target individual with the previously determined reporting order, the reporter whose speaking order matches the target individual's is identified. This reporter is the one corresponding to the target individual's speaking order. By analyzing the reporting methods used to present information, the identity of the target individual can be reasonably inferred based on the reporter's identity information under different reporting methods. This improves the efficiency of target individual identity recognition while ensuring its accuracy.

[0087] 308. The computer device performs object detection on the second image to determine the target region, which is the region where the target object is located in the second image.

[0088] In this embodiment, the second image refers to the image captured by the camera during the meeting, which includes information such as all participants in the meeting, the environmental setup, and any items that may identify them. After acquiring the second image, the computer device can use an object detection algorithm to perform object detection on the second image to determine the specific location of the target object in the second image, thereby delineating the target area where the target object is located.

[0089] Optionally, during object detection, the computer device can locate the target object in the second image based on lip movements. Specifically, after acquiring the second image captured by the camera, the computer device can treat it as a series of continuous image frames and analyze these frames one by one to track the movement of each pixel in the image. Since the lips of the target object will have obvious shape changes and movements when speaking in a meeting scenario, by analyzing the movement trajectory and change characteristics of the pixels in the lip region of the second image, the target object among the multiple attendees included in the second image can be identified. In other words, the computer device can identify attendees whose lips are in motion in the second image as the target object.

[0090] 309. The computer device performs at least one of face recognition and text recognition on the second image based on the target area to obtain the identity information of the target object.

[0091] In this embodiment, since the target area includes not only the facial information of the target object but also text information related to the target object, the computer device can perform at least one of facial recognition and text recognition on the target area to obtain the identity information of the target object. By first determining the identity information of the target object from the image area where the target object is located in the second image, and then performing at least one of facial recognition and text recognition on that image area, compared to performing the above recognition on the entire second image, the scope of facial recognition and text recognition can be greatly reduced, thereby further improving the efficiency of target object identification.

[0092] In some embodiments, the computer device may choose to use facial recognition or text recognition alone, or it may use both technologies simultaneously to obtain the identity information of the target object, depending on the specific circumstances. Accordingly, step 309 above includes the following three implementation methods:

[0093] Implementation Method 1: The computer device performs face recognition on the second image based on the target area to obtain the face recognition result of the target object. The face recognition result is used to indicate whether the target object's face has been registered. If the face recognition result indicates that the target object's face has been registered, the identity information corresponding to the face is determined as the target object's identity information. Face recognition involves analyzing the target object's facial features and comparing them with existing face templates in a face database to determine if a matching face template exists in the database. Optionally, this face database is the face database associated with the access control system of the target object's company. If a face template matching the target object's face exists in the face database, the computer device can determine the identity information corresponding to the matching face template as the target object's identity information. By recognizing that the target object's face has been registered and determining the identity information corresponding to that registered face as the target object's identity information, the accuracy of target object identification can be improved.

[0094] Method Two: The computer device performs text recognition on the second image based on the target area to obtain the text recognition result of the target object. The text recognition result indicates whether the target area contains text information indicating identity. If the text recognition result indicates that the target area contains text information indicating identity, the text information is identified as the target object's identity information. Specifically, when text recognition is selected, the computer device focuses on documents that may indicate identity within the target area, such as name tags, hanging name tags, employee badges, or name badges. Further image enhancement processing is applied to these areas to highlight the text for better recognition. Then, OCR technology is used to recognize the text in these areas, converting the text in the image into text. Afterward, key identity-related information, such as name and position, is extracted from the recognized text. For example, recognizing "Name: Zhang San, Position: Project Manager" from a name tag confirms the target object's identity information. By recognizing identity-related text information in the image area where the target object is located, and identifying that text information as the target object's identity information, the accuracy of target object identification in this meeting can be further improved.

[0095] Implementation Method 3: The computer device performs facial recognition and text recognition on the second image based on the target area, obtaining the facial recognition result and text recognition result of the target object. The facial recognition result indicates whether the target object's face has been registered, and the text recognition result indicates whether the target area includes text information indicating identity. Based on the facial recognition result and text recognition result, the identity information of the target object is determined. Specifically, if the facial recognition result indicates that the target object's face is registered and the text recognition result indicates that the target area includes text information indicating identity, the computer device can match and verify these two results. If the facial recognition results show the target object as "Li Ming, Marketing Department," and the text recognition result shows "Li Ming, Marketing Department Manager" from the identity card, the two results corroborate each other in terms of name and department information, further confirming the accuracy of the identity information. At this point, combining the two information, the identity information of the target object is determined to be "Li Ming, Marketing Department Manager." If the facial recognition result indicates that the target object's face is registered and the text recognition result indicates that the target area does not include text information indicating identity, the computer device can prioritize the facial recognition result to determine the target object's identity information. When facial recognition results indicate that the target's face is not registered, and text recognition results indicate that the target area includes text information used to represent identity, the computer device can determine the text information as the target's identity information. By jointly determining the target's identity information based on both facial recognition and text recognition results, the accuracy of target identification can be further improved.

[0096] It should be noted that, in addition to the schemes for identifying the target object based on the voice of other participants in steps 303-304, the schemes for identifying the target object based on the first image displayed on the screen in the meeting in steps 305-307, and the schemes for identifying the target object based on the second image captured by the camera in the meeting in steps 308-309, the computer device can also randomly combine the above three identification schemes. For example, based on the identification scheme in steps 303-304, it can combine at least one of the identification schemes in steps 305-307 and steps 308-309. That is, the embodiments of this application also protect the schemes for identifying the target object based on the voice and the first image of other participants, the schemes for identifying the target object based on the voice and the second image of other participants, and the schemes for identifying the target object based on the voice, the first image, and the second image of other participants.

[0097] In some embodiments, after determining the identity information of the target object, the computer device can also record the target object's voiceprint and the corresponding identity information into a voiceprint database. By recording the target object's voiceprint and the corresponding identity information into a voiceprint database after determining the target object's identity information, automated voiceprint recording can be achieved, eliminating the need for users to actively record their voiceprints and improving the efficiency of human-computer interaction.

[0098] The following is through Figure 4 The flowchart shown illustrates the overall identity recognition process, providing a general explanation of the identity recognition scheme. Figure 4 As shown, the computer device can acquire the voice of the current target person in the meeting and perform voiceprint recognition. If a voiceprint template matching the target person's voiceprint is identified from the voiceprint database, it indicates that the voiceprint database stores the target person's identity information, meaning the target person actively registered their voiceprint before the meeting. Therefore, the computer device can output the target person's identity information. Conversely, if no voiceprint template matching the target person's voiceprint is identified from the voiceprint database, the computer device will enter the multimodal speaker identification module. The multimodal speaker automatic identification module includes a speech content analysis module (corresponding to...). Figure 4 Module 1) Screen Image Analysis Module (corresponding to) Figure 4 Module 2) and the camera image analysis module (corresponding to Figure 4 Module 3 in the document. The speech content analysis module analyzes the speech of participants other than the target audience to identify the target audience. See also... Figure 5 The diagram shown illustrates the speech content analysis module, which includes speech recognition (corresponding to...). Figure 4 Speech content analysis and text content analysis (corresponding to) Figure 4The process involves two steps: context analysis and dialogue analysis. Text content analysis specifically includes proximity role analysis, presentation order analysis, and dialogue analysis. Proximity role analysis determines the target's identity based on the voices of attendees who speak before the target. Presentation order analysis determines the target's identity based on the order in which other attendees speak. Dialogue analysis determines the target's identity based on the voices of attendees who speak after the target. The screen image analysis module analyzes the images displayed on the screen to identify the target, including screen text analysis and presenter (PPT speaker) analysis. The camera image analysis module analyzes images captured by the camera to identify the target, including image content analysis, text recognition (corresponding to attendee analysis in the image), or facial recognition. After identifying the target through at least one of the above modules, the computer can also update the voiceprint database, recording the target's voiceprint and its corresponding voiceprint into the database.

[0099] This application provides an identity recognition method that can identify the target object in a meeting based on the meeting information when the identity of the target object cannot be identified by voiceprint. This eliminates the need for the user to actively record their voiceprint, thus achieving automatic identification of the user's identity and greatly improving the efficiency of identity recognition.

[0100] Figure 6 This is a schematic diagram of the structure of an identity recognition device provided in an embodiment of this application. See also... Figure 6 The device includes: a first acquisition module 601, a second acquisition module 602, and an identity recognition module 603.

[0101] The first acquisition module 601 is used to acquire the voiceprint recognition result of the target object based on the voice of the target object in the meeting. The voiceprint recognition result is used to indicate whether the voiceprint of the target object has been registered.

[0102] The second acquisition module 602 is used to acquire meeting information when the voiceprint recognition result indicates that the voiceprint of the target object is not registered;

[0103] The identity recognition module 603 is used to identify the target object based on meeting information.

[0104] In some embodiments, the meeting information includes the voices of other participants in the meeting besides the target. Figure 7 This is a schematic diagram of another identity recognition device provided in an embodiment of this application. See also... Figure 7 The identity recognition module 603 includes:

[0105] The recognition unit 701 is used to perform speech recognition on the speech of other participants to obtain the speech information of other participants besides the target object;

[0106] The determining unit 702 is used to determine the speech information related to the target object from the speech information of other participants besides the target object;

[0107] The determining unit 702 is also used to determine the identity information of the target object based on the speech information related to the target object.

[0108] In some embodiments, the determining unit 702 is configured to determine, from the speaking information of other participants besides the target object, the speaking information within a first preset time period before the target object speaks; determine the identity information of the next participant to speak from the speaking information within the first preset time period before the target object speaks; and determine the identity information of the next participant to speak as the identity information of the target object.

[0109] In some embodiments, the determining unit 702 is configured to determine, from the speaking information of other participants besides the target object, speaking information related to the speaking order of the meeting; determine, from the speaking information related to the speaking order of the meeting, the identity information of the participants corresponding to the speaking order of the target object; and determine the identity information of the participants as the identity information of the target object.

[0110] In some embodiments, the determining unit 702 is configured to determine, from the speaking information of other participants besides the target object, the speaking information within a second preset time period after the target object speaks; determine the identity information of the previous speaker from the speaking information within the second preset time period after the target object speaks; and determine the identity information of the previous speaker as the identity information of the target object.

[0111] In some embodiments, the meeting information includes a first image displayed on the screen during the meeting. The first image includes the meeting's display information. The identity recognition module 603 is used to perform text recognition on the first image to obtain the meeting's display information; determine the identity information of the presenter of the display information from the display information; and determine the identity information of the target object based on the identity information of the presenter of the display information.

[0112] In some embodiments, the identity recognition module 603 is used to determine the identity information of the reporter as the identity information of the target object when the reporter of the displayed information is a single person and the speech information of the target object is related to the displayed information; when the reporter of the displayed information is multiple people, the reporter corresponding to the speech information or speech order of the target object is determined based on the reporting order of the multiple reporters and the displayed information corresponding to each reporter; and the identity information of the reporter corresponding to the speech information or speech order of the target object is determined as the identity information of the target object.

[0113] In some embodiments, the meeting information includes a second image captured by a camera during the meeting. The second image includes all participants in the meeting. The identity recognition module 603 is used to perform object detection on the second image, determine the target area, which is the area where the target object is located in the second image. Based on the target area, at least one of face recognition and text recognition is performed on the second image to obtain the identity information of the target object.

[0114] In some embodiments, the identity recognition module 603 is used to perform face recognition on the second image based on the target region to obtain the face recognition result of the target object. The face recognition result is used to indicate whether the face of the target object has been registered. If the face recognition result indicates that the face of the target object has been registered, the identity information corresponding to the face is determined as the identity information of the target object.

[0115] In some embodiments, the identity recognition module 603 is used to perform text recognition on the second image based on the target region to obtain the text recognition result of the target object. The text recognition result is used to indicate whether the target region includes text information used to represent identity. If the text recognition result indicates that the target region includes text information used to represent identity, the text information is determined as the identity information of the target object.

[0116] In some embodiments, the identity recognition module 603 is used to perform face recognition and text recognition on the second image based on the target area to obtain the face recognition result and text recognition result of the target object. The face recognition result is used to indicate whether the face of the target object has been registered, and the text recognition result is used to indicate whether the target area includes text information used to represent identity. Based on the face recognition result and text recognition result, the identity information of the target object is determined.

[0117] In some embodiments, see continue to see Figure 7 The device further includes:

[0118] The input module 604 is used to input the voiceprint of the target object and the corresponding identity information into the voiceprint database.

[0119] This application provides a target object identification device that can identify the target object based on the meeting information when the identity of the target object in the meeting cannot be identified by voiceprint. This eliminates the need for the user to actively record their voiceprint, thus achieving automatic identification of the user's identity and greatly improving the efficiency of identification.

[0120] It should be noted that the identity recognition device provided in the above embodiments is only an example of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above. In addition, the identity recognition device and the identity recognition method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0121] This application also provides a computer device, which includes a processor and a memory. The memory stores at least one computer program, which is loaded and executed by the processor to implement the identity recognition method described in the above embodiments.

[0122] Figure 8 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application.

[0123] The computer device 800 can vary considerably due to differences in configuration or performance. It may include one or more Central Processing Units (CPUs) 801 and one or more memories 802. The memory 802 stores at least one computer program, which is loaded and executed by the processor 801 to implement the identity recognition methods provided in the various method embodiments described above. Of course, the computer device 800 may also have wired or wireless network interfaces, a keyboard, and input / output interfaces for input and output. The computer device 800 may also include other components for implementing device functions, which will not be elaborated upon here.

[0124] This application also provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to implement the identity recognition method described above.

[0125] This application also provides a computer program product, including a computer program that is loaded and executed by a processor to implement the identity recognition method as described in the above embodiments.

[0126] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0127] The above are merely optional embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present application should be included within the protection scope of the present application.

Claims

1. An identity recognition method, characterized in that, The method includes: Based on the speech of the target object in the meeting, the voiceprint recognition result of the target object is obtained, and the voiceprint recognition result is used to indicate whether the voiceprint of the target object has been registered; If the voiceprint recognition result indicates that the target object's voiceprint is not registered, obtain the meeting information of the meeting; Based on the meeting information, the target object is identified.

2. The method according to claim 1, characterized in that, The meeting information includes the voices of other participants in the meeting besides the target individual. The step of identifying the target individual based on the meeting information includes: Speech recognition is performed on the voices of the other participants to obtain the speaking information of the other participants besides the target object; From the speech information of other participants besides the target, determine the speech information related to the target; Based on the speech information related to the target object, the identity information of the target object is determined.

3. The method according to claim 2, characterized in that, Determining the speech information related to the target from the speech information of other participants besides the target includes: From the speaking information of other participants besides the target, determine the speaking information within a first preset time period before the target speaks; The step of determining the identity information of the target object based on speech information related to the target object includes: The identity information of the next speaker is determined from the speech information within a first preset time period before the target person speaks. The identity information of the next speaker is determined as the identity information of the target object.

4. The method according to claim 2, characterized in that, Determining the speech information related to the target from the speech information of other participants besides the target includes: From the speaking information of other participants besides the target, determine the speaking information related to the speaking order of the meeting; The step of determining the identity information of the target object based on speech information related to the target object includes: From the speaking information related to the speaking order of the meeting, determine the identity information of the participants corresponding to the speaking order of the target object; The identity information of the attendees is determined as the identity information of the target object.

5. The method according to claim 2, characterized in that, Determining the speech information related to the target from the speech information of other participants besides the target includes: From the speaking information of other participants besides the target, determine the speaking information within a second preset time period after the target speaks; The step of determining the identity information of the target object based on speech information related to the target object includes: The identity information of the previous speaker is determined from the speech information within a second preset time period after the target person speaks. The identity information of the participant who spoke last is determined as the identity information of the target object.

6. The method according to claim 1, characterized in that, The meeting information includes a first image displayed on the screen during the meeting, the first image including the meeting's display information. The step of identifying the target object based on the meeting information includes: Perform text recognition on the first image to obtain the display information of the meeting; From the displayed information, determine the identity information of the person presenting the information; Based on the identity information of the presenter of the displayed information, the identity information of the target object is determined.

7. The method according to claim 6, characterized in that, The process of determining the identity information of the target object based on the identity information of the presenter of the displayed information includes: If the presenter of the displayed information is a single person and the speech information of the target object is related to the displayed information, the identity information of the presenter shall be determined as the identity information of the target object. When there are multiple presenters for the information being displayed, the presenter corresponding to the speaking information or speaking order of the target object is determined based on the speaking order of the multiple presenters and the display information corresponding to each presenter. The identity information of the reporter corresponding to the speech information or speech order of the target object is determined as the identity information of the target object.

8. The method according to claim 1, characterized in that, The meeting information includes a second image captured by a camera during the meeting, the second image including all participants in the meeting. The step of identifying the target based on the meeting information includes: Object detection is performed on the second image to determine the target region, which is the region where the target object is located in the second image; Based on the target region, at least one of face recognition and text recognition is performed on the second image to obtain the identity information of the target object.

9. The method according to claim 8, characterized in that, The step of performing at least one of face recognition and text recognition on the second image based on the target region to obtain the identity information of the target object includes: Based on the target region, face recognition is performed on the second image to obtain the face recognition result of the target object. The face recognition result is used to indicate whether the face of the target object has been registered. If the face recognition result indicates that the target object's face has been registered, the identity information corresponding to the face is determined as the target object's identity information.

10. The method according to claim 8, characterized in that, The step of performing at least one of face recognition and text recognition on the second image based on the target region to obtain the identity information of the target object includes: Based on the target region, text recognition is performed on the second image to obtain the text recognition result of the target object. The text recognition result is used to indicate whether the target region includes text information used to represent identity. If the text recognition result indicates that the target area includes text information used to represent identity, the text information is identified as the identity information of the target object.

11. The method according to claim 8, characterized in that, The step of performing at least one of face recognition and text recognition on the second image based on the target region to obtain the identity information of the target object includes: Based on the target region, face recognition and text recognition are performed on the second image to obtain the face recognition result and text recognition result of the target object. The face recognition result is used to indicate whether the face of the target object has been registered, and the text recognition result is used to indicate whether the target region includes text information used to indicate identity. Based on the facial recognition results and the text recognition results, the identity information of the target object is determined.

12. The method according to claim 1, characterized in that, The method further includes: The voiceprint of the target object and the corresponding identity information are entered into the voiceprint database.

13. An identity recognition device, characterized in that, The device includes: The first acquisition module is used to acquire the voiceprint recognition result of the target object based on the voice of the target object in the meeting, and the voiceprint recognition result is used to indicate whether the voiceprint of the target object has been registered. The second acquisition module is used to acquire meeting information of the meeting when the voiceprint recognition result indicates that the voiceprint of the target object is not registered; The identity recognition module is used to identify the target object based on the meeting information.

14. A computer device, characterized in that, The computer device includes a processor and a memory, the memory storing at least one computer program, which is loaded and executed by the processor to implement the identity recognition method as described in any one of claims 1 to 12.

15. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, which is loaded and executed by a processor to implement the identity recognition method as described in any one of claims 1 to 12.

Images