Face recognition method, priority determination method, device and storage medium

Abstract

The present application provides a face recognition method, a priority determination method, an apparatus, and a storage medium. The method comprises: obtaining an image to be recognized for face recognition using a face recognition application; wherein the image to be recognized includes a face image of a user to be recognized; based on the priorities of multiple user databases, preferentially utilizing a user database with a relatively high priority among the multiple user databases to perform face recognition on the image to be recognized; wherein the user databases store face images of users; and the priority of each user database is determined based on the probability that the user corresponding to the face image stored in each user database will use the face recognition application, and the priority of the user database is positively correlated with the magnitude of the probability. In this way, for face databases with a large data size, face recognition operations are preferentially performed on user databases with a high probability, which can improve the accuracy of the recognition results.

Description

Technical Field

[0001] The present application relates to the field of face recognition technology, and in particular to a face recognition method, a priority determination method, a device, and a storage medium. Background Art

[0002] In the related art, face recognition technology is based on the convolutional neural network algorithm, which improves recognition accuracy by using a large number of public data sets for training and designing more appropriate convolutional neural network structures. At present, the accuracy of face recognition technology has achieved very good results in the application of small-scale face databases (or called face base databases), and in some application scenarios, the accuracy can even reach more than 99%. However, in the application scenarios of large-scale face base databases (the number of faces reaches millions or even tens of millions), the accuracy of face recognition technology in the related art drops significantly. This is because when the scale of the face database rises to the tens of millions level, the existing convolutional neural network structure is not able to calculate particularly ideal facial feature vector values, so when searching in a large-scale face database, the accuracy drops significantly. Summary of the Invention

[0003] The present application provides a face recognition method, a priority determination method, an apparatus, and a storage medium. When faced with a large-scale face database, retrieval is performed preferentially among high-probability groups based on the probability of users accessing the face recognition application, thereby improving the accuracy of the face recognition results.

[0004] The technical solution of the embodiment of the present application is implemented as follows:

[0005] In a first aspect, an embodiment of the present application provides a face recognition method, the method comprising:

[0006] Acquire an image to be recognized for performing a face recognition operation using a face recognition application; wherein the image to be recognized includes a face image of a user to be recognized;

[0007] Based on the priorities of multiple user databases, a user database with a relatively higher priority among the multiple user databases is preferentially used to perform a face recognition operation on the image to be recognized; wherein, the user database stores the face images of users; the priority of each user database is determined according to the probability that the user corresponding to the face image stored in each user database uses the face recognition application, and the priority of the user database is positively correlated with the size of the probability.

[0008] In a second aspect, an embodiment of the present application provides a priority determination method, the method comprising:

[0009] Obtain facial images of multiple users;

[0010] Calculating probabilities that the multiple users use the face recognition application;

[0011] Based on the facial images of the multiple users and the probabilities of the multiple users using the facial recognition application, multiple user databases are established; wherein the priority of each of the multiple user databases is determined according to the probability of the users corresponding to the facial images stored in each user database using the facial recognition application; the priority of the user database is positively correlated with the size of the probability.

[0012] In a third aspect, an embodiment of the present application provides a face recognition device, which includes a receiving unit and a recognition unit, wherein:

[0013] A receiving unit configured to obtain an image to be recognized for performing a face recognition operation using a face recognition application; wherein the image to be recognized includes a face image of a user to be recognized;

[0014] The recognition unit is configured to perform a face recognition operation on the image to be recognized by preferentially utilizing a user database with a relatively higher priority among the multiple user databases based on the priorities of the multiple user databases; wherein the user database stores facial images of users; the priority of each user database is determined according to the probability that the user corresponding to the facial image stored in each user database uses the face recognition application, and the priority of the user database is positively correlated with the size of the probability.

[0015] In a fourth aspect, an embodiment of the present application provides a face recognition device, which includes a memory and a processor; wherein,

[0016] The memory is used to store a computer program that can be run on the processor;

[0017] The processor is configured to execute the steps of the method described in the first aspect when running the computer program.

[0018] In a fifth aspect, an embodiment of the present application provides a priority determination device, which includes an image unit, a calculation unit, and a determination unit, wherein:

[0019] an imaging unit configured to obtain facial images of a plurality of users;

[0020] a calculation unit configured to calculate a probability that a plurality of users use a face recognition application;

[0021] A determination unit is configured to establish multiple user databases based on the facial images of the multiple users and the probability of the multiple users using the facial recognition application; wherein the priority of each of the multiple user databases is determined according to the probability of the users corresponding to the facial images stored in each user database using the facial recognition application; and the priority of the user database is positively correlated with the size of the probability.

[0022] In a sixth aspect, an embodiment of the present application provides a priority determination device, which includes a memory and a processor; wherein,

[0023] The memory is used to store a computer program that can be run on the processor;

[0024] The processor is used to execute the steps of the method described in the second aspect when running the computer program.

[0025] In a seventh aspect, an embodiment of the present application provides a computer storage medium storing a face recognition program or a priority determination program; wherein,

[0026] When the face recognition program is executed by at least one processor, the steps of the method according to the first aspect are implemented;

[0027] When the priority determination program is executed by at least one processor, the steps of the method according to the second aspect are implemented.

[0028] In an eighth aspect, the present application provides a face recognition system, which at least includes the face recognition device as described in the third aspect or the fourth aspect, and the priority determination device as described in the fifth aspect or the sixth aspect.

[0029] The present application provides a face recognition method, a priority determination method, an apparatus, and a storage medium. The method comprises: obtaining an image to be recognized for face recognition using a face recognition application; wherein the image to be recognized includes a face image of a user to be recognized; based on the priorities of multiple user databases, preferentially utilizing a user database with a relatively high priority among the multiple user databases to perform face recognition on the image to be recognized; wherein the user databases store face images of users; and wherein the priority of each user database is determined based on the probability that the user corresponding to the face image stored in each user database will use the face recognition application, and the priority of the user database is positively correlated with the magnitude of the probability. In this way, for a face database with a large data size, it can be divided into multiple user databases according to the probability of users using the face recognition application, and face recognition operations can be preferentially performed on user databases with a high usage probability, thereby improving the accuracy of the recognition results. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A process diagram of a face recognition method provided in the related art;

[0031] Figure 2 A flowchart of a face recognition method provided in an embodiment of the present application;

[0032] Figure 3 A flowchart of a priority determination method provided in an embodiment of the present application;

[0033] Figure 4 Examples of different facial images of the same user provided in the embodiments of this application;

[0034] Figure 5 A flowchart of another face recognition method provided in an embodiment of the present application;

[0035] Figure 6 A flowchart of another face recognition method provided in an embodiment of the present application;

[0036] Figure 7 A schematic diagram of the structure of a face recognition device provided in an embodiment of the present application;

[0037] Figure 8 A schematic diagram of the specific hardware structure of a face recognition device provided in an embodiment of the present application;

[0038] Figure 9 A schematic diagram of the structure of a priority determination device provided in an embodiment of the present application;

[0039] Figure 10 A schematic diagram of the specific hardware structure of a priority determination device provided in an embodiment of the present application;

[0040] Figure 11 A schematic diagram of the structure of a face recognition system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0041] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application.

[0042] In the field of face recognition, the commonly used method is to use a large amount of face image data to train a convolutional neural network. Figure 1 , which shows a process diagram of a face recognition method provided in the related art. Figure 1As shown, a convolutional neural network is used to perform calculations on face images 1, 2, ..., n, to obtain facial feature vectors 1, 2, ..., n, which represent facial features. In other words, each face image can be used as input to a trained convolutional neural network. Through a series of calculations, the convolutional neural network ultimately generates a facial feature vector of a fixed length (for example, 128 dimensions, 512 dimensions, or even more). Generally, facial feature vectors for the same person are highly similar, while facial feature vectors for different people are less similar. Therefore, this method can be used for face recognition.

[0043] The accuracy of existing face recognition has achieved good results in applications with small-scale face databases. By using large public data sets for training and designing more appropriate convolutional neural network structures, existing face recognition methods have achieved good results, and in some application scenarios, the accuracy can even reach over 99%. However, in application scenarios with large-scale face databases (the number of faces reaches millions or even tens of millions), the accuracy of existing face recognition methods has dropped significantly. This is mainly due to two reasons. First, there is a deviation between the face image data used for training and the face image data collected for actual applications. Second, when the size of the face database increases to the tens of millions, the existing convolutional neural network structure is not able to calculate particularly ideal facial feature vector values.

[0044] Based on this, an embodiment of the present application provides a face recognition method, the basic concept of which is as follows: obtaining an image to be recognized for face recognition using a face recognition application; wherein the image to be recognized includes a face image of a user to be recognized; based on the priorities of multiple user databases, preferentially utilizing a user database with a relatively high priority among the multiple user databases to perform face recognition on the image to be recognized; wherein the user database stores face images of users; the priority of each user database is determined based on the probability that the user corresponding to the face image stored in each user database will use the face recognition application, and the priority of the user database is positively correlated with the magnitude of the probability. In this way, for a face database with a large data size, it can be divided into multiple user databases according to the probability of users using the face recognition application, and face recognition operations are preferentially performed on user databases with a high usage probability, which can improve the accuracy of the recognition results.

[0045] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0046] In one embodiment of the present application, see Figure 2 , which shows a flow chart of a face recognition method provided by an embodiment of the present application, such as Figure 2As shown, the method may include:

[0047] S101: Acquire an image to be recognized for performing a face recognition operation using a face recognition application.

[0048] It should be noted that the execution entity of the face recognition method provided in the embodiments of the present application can be a server, a face recognition application, or a face recognition device installed with a face recognition application. Among them, the face recognition application can be set on a terminal device with a relatively fixed position, such as an ATM, a supermarket self-service cash register, an access control device, etc.; in addition, the face recognition application can also be set on a terminal device with a movable position, such as a user equipment (UE), a personal digital assistant (PDA), a handheld device, a computing device, a vehicle-mounted device, a wearable device, etc. The face recognition application can be a payment application, a communication chat application, an audio and video application, or other applications that require a face recognition function.

[0049] It should also be noted that the image to be recognized includes a facial image of the user to be recognized. To perform facial recognition, an image to be recognized must be acquired, and the image to be recognized must include the facial image of the user to be recognized. In other words, the image to be recognized only needs to include a portion of the user's face. The image to be recognized can be captured by the facial recognition device using a camera, or it can be obtained from another device via wireless or wired transmission.

[0050] S102: Based on the priorities of the multiple user databases, preferentially utilizing a user database with a relatively higher priority among the multiple user databases to perform a face recognition operation on the image to be recognized.

[0051] It should be noted that the user database stores user facial images. The user database serves as a blueprint for facial recognition, storing multiple user facial images and the corresponding identity information. The facial recognition process involves determining the identity information corresponding to the image to be recognized based on the multiple user facial images stored in the user database. Furthermore, multiple user databases are established based on user data. This user data can be obtained from historical base station access information, or it can be the original user database already established by the facial recognition application, which is simply divided based on the base station's historical access information to generate multiple user databases.

[0052] Secondly, the priority of each user database is determined according to the probability that the user corresponding to the facial image stored in each user database uses the facial recognition application (hereinafter referred to as probability or application usage probability), and the priority of the user database is positively correlated with the size of the probability. In an embodiment of the present application, there are multiple user databases, and the user groups included in each user database are different. At the same time, the multiple user databases are prioritized according to the probability of users in the user database using the facial recognition application. That is to say, if the probability of users in the user database using the facial application is very high, the priority of the user database is high, and the user database will be used first to perform facial recognition operations on the image to be identified. In this way, the image to be identified can be retrieved from the high-probability population first. If it cannot be retrieved, it will be searched from the low-probability population, thereby improving the accuracy of facial recognition. In this way, the processing volume during the facial recognition operation can be reduced, and the accuracy of the facial recognition results can also be guaranteed.

[0053] It should also be noted that the application usage probability can be calculated using a variety of algorithms, so it can be calculated based on various types of data. For example, the application usage probability can be determined based on historical data of the face recognition application, it can be determined based on relevant data of the deployment location of the face recognition application, it can be determined based on historical data of the device corresponding to the face recognition application, and so on.

[0054] Furthermore, in some embodiments, the application usage probability is determined based on specified statistical data of a mobile communication network operator that provides mobile data services for the face recognition application;

[0055] The specified statistical data can represent the size of the probability.

[0056] It should be noted that in the embodiments of the present application, the aforementioned probability is determined based on the specified statistical data of the mobile communication network operator that provides mobile data services for the facial recognition application. Mobile communication network operators have access to a large amount of user data, such as the user's geographic location each time they access the mobile communication network and the time they stay at that location. Therefore, the specified statistical data of the mobile communication network operator can be used to calculate the probability of the user using the facial recognition application.

[0057] The specified statistical data can represent the size of the application usage probability. It should be understood that the specific content of the specified statistical data needs to be determined based on the calculation method of the application usage probability in the application scenario. For example, the specified statistical data may include the geographical location of the user each time they access the mobile communication network, the access duration, etc. This application does not limit the specific type of the specified statistical data.

[0058] Furthermore, in some embodiments, the specified statistical data may include base station distance data and service probability data; wherein,

[0059] The base station distance data represents the distance between the face recognition application and a base station capable of providing a mobile data service for the face recognition application; and

[0060] The service probability data indicates the probability that the base station provides the mobile data service for the user.

[0061] It should be noted that if the aforementioned probability is the probability of a user accessing the location where the facial recognition application is deployed, then the designated statistical data may include base station distance data and service probability data. In this case, the base station distance data represents the distance between the facial recognition application and multiple base stations that can provide mobile data services for the facial recognition application, and the service probability data represents the probability that the multiple base stations will provide mobile data services to the user. Using the base station distance data and service probability data, the probability corresponding to each user can be calculated, thereby preferentially utilizing the user database with a relatively high priority among the multiple user databases to perform facial recognition operations on the image to be recognized.

[0062] Since the facial images stored in the user database and the images to be recognized are unlikely to be taken in the same environment, which affects the accuracy of the recognition process, in some embodiments, the facial images of users stored in the user database include: a high-quality facial image set consisting of high-quality facial images, and a low-quality facial image set consisting of low-quality facial images;

[0063] The high-quality facial image refers to the facial image retained when the user handles a business with relatively high requirements for facial image quality; the low-quality facial image refers to the facial image retained when the user handles a business with relatively low requirements for facial image quality.

[0064] It should be noted that since the shooting background has a significant impact on the accuracy of facial recognition operations, using facial images from multiple shooting environments to perform comprehensive facial recognition operations can improve recognition accuracy. Specifically, the user database can store a high-quality facial image set consisting of high-quality facial images and a low-quality facial image set consisting of low-quality facial images. Here, high-quality facial images refer to facial images retained when users handle services that require relatively high facial image quality; low-quality facial images refer to facial images retained when users handle services that require relatively low facial image quality.

[0065] That is to say, for any specific user in the user database, the user's identity information, high-quality face image set, low-quality face image set and other information are stored.

[0066] Accordingly, in some embodiments, the step of preferentially utilizing a user database having a relatively higher priority among the multiple user databases to perform the face recognition operation on the image to be recognized may include:

[0067] Determining a user database that is preferentially used for performing a face recognition operation on the image to be recognized;

[0068] Based on the determined set of high-quality facial images and the set of low-quality facial images in the user database, performing facial recognition on the images to be recognized, respectively, to obtain a first facial similarity judgment result corresponding to the set of high-quality facial images and a second facial similarity judgment result corresponding to the set of low-quality facial images;

[0069] The face recognition result of the image to be recognized is obtained by fusing the first face similarity judgment result and the second face similarity judgment result.

[0070] It should be noted that when performing a face recognition operation, first, a user database is determined from multiple user databases to be preferentially used for face recognition operations on the image to be recognized. Then, a face comparison is performed with the user using a set of high-quality face images and a set of low-quality face images in the user database, thereby obtaining a first face similarity judgment result corresponding to the set of high-quality face images and a second face similarity judgment result corresponding to the set of low-quality face images. Finally, the first face similarity judgment result and the second face similarity judgment result are combined to obtain a face recognition result for the image to be recognized. In this way, the influence of the shooting environment on the face image can be reduced, thereby improving the accuracy of the face recognition operation.

[0071] In a specific embodiment, the user database includes N sub-databases; N is a positive integer greater than or equal to 2.

[0072] The i-th sub-database includes the i-th facial images of multiple users in the user database, and i is a positive integer less than or equal to N.

[0073] It should be noted that in the embodiment of the present application, the user database is composed of N sub-databases, and the i-th sub-database includes the i-th facial image of multiple users in the user database. In other words, for each user in the user database, N images are stored. These N images can be user facial images from different sources and different shooting scenes. These N images constitute the aforementioned high-quality facial image set and low-quality facial image set. Then, the i-th image of all users constitutes the i-th sub-database. Exemplarily, the user database stores four different images for each user. Then, the first facial image of all users constitutes the first sub-database, the second facial image of all users constitutes the second sub-database, the third facial image of all users constitutes the third sub-database, and the fourth facial image of all users constitutes the fourth sub-database.

[0074] Since the user database contains multiple sub-databases, the face recognition operation is also performed based on the multiple sub-databases. Therefore, in some embodiments, the face recognition operation on the image to be recognized in the high-priority user database may specifically include:

[0075] Based on the N sub-databases corresponding to the high-priority user database, performing face recognition operations on the image to be identified respectively, and obtaining N sub-similarity between the face image of each user in the high-priority user database and the face image of the user to be identified;

[0076] Determining a target similarity between the facial image of each user and the facial image of the user to be identified based on the obtained N sub-similarity scores;

[0077] Based on the determined target similarity, it is determined whether there is a target user meeting a preset standard in the high-priority user database.

[0078] It should be noted that for the high-priority user database, the face recognition operation specifically includes the following steps: (1) comparing the similarity of the image to be identified with the face images of each user in the N sub-databases to obtain N sub-similarity scores corresponding to each user; (2) obtaining the target similarity score corresponding to each user by summing, weighted summing, or averaging the N sub-similarity scores of each user. In this way, based on the target similarity score corresponding to each user, it is possible to determine whether the user is the target user, thereby determining the identity information corresponding to the image to be identified. In addition, if the user to be identified cannot be identified in the high-priority user database, it can be identified in the next-priority user database, and the identification process is the same as above.

[0079] In a specific example, N is equal to 4, and determining the target similarity between the facial image of each user and the facial image of the user to be identified may include:

[0080] Based on the first sub-database corresponding to the high-priority user database, matching the image to be identified with the first face image of each user in the first sub-database to obtain a first sub-similarity of each user;

[0081] Comparing the first sub-similarity of each user with a preset threshold to determine a plurality of pre-identified users;

[0082] Based on the second sub-database corresponding to the high-priority user database, matching the image to be identified with the second facial images of the plurality of pre-identified users in the second sub-database, respectively, to obtain a second sub-similarity for each pre-identified user;

[0083] Based on the third sub-database corresponding to the high-priority user database, matching the image to be identified with the third facial images of the plurality of pre-identified users in the third sub-database, respectively, to obtain a third sub-similarity for each pre-identified user;

[0084] Based on the fourth sub-database corresponding to the high-priority user database, matching the image to be identified with the fourth facial images of the plurality of pre-identified users in the fourth sub-database, respectively, to obtain a fourth sub-similarity for each pre-identified user;

[0085] A weighted sum of the first sub-similarity, the second sub-similarity, the third sub-similarity, and the fourth sub-similarity corresponding to each pre-identified user is calculated, and the obtained weighted sum is determined as the target similarity between the facial image of each pre-identified user and the facial image of the user to be identified.

[0086] It should be noted that, in a specific application scenario, the user database includes four different sub-databases, wherein the first sub-database stores the first face image of each user, the second sub-database stores the second face image of each user, the third sub-database stores the third face image of each user, and the fourth sub-database stores the fourth face image of each user. Figure 4 , which shows examples of different facial images of the same user provided by the embodiment of the present application. Figure 4 As shown in , the images of the same user in different scenes are slightly different. Figure 4 As shown in a, the first face image can be a photo of the user who handles the business on site; Figure 4 As shown in b, the second face image can be a photo of the user who handles the business on site; Figure 4As shown in c, the third face image can be a user photo captured by a camera; Figure 4 As shown in FIG. 4 , the fourth facial image may be an ID photo uploaded by the user when handling a business or an ID photo collected using a scanner or other equipment.

[0087] It should also be noted that the specific operations of face recognition based on the four sub-databases are as follows:

[0088] (1) In the first sub-database, the first face image of each user is compared with the image to be identified for similarity, and the first sub-similarity corresponding to each user is obtained. Users whose first sub-similarity is greater than or equal to a preset threshold are determined as pre-identified users. In this case, there are multiple pre-identified users. In this way, the pre-identification operation can exclude most users who are completely different from the image to be identified, thereby reducing the subsequent processing volume. The preset threshold can be determined according to the specific usage scenario.

[0089] (2) in the second sub-database, by comparing the second facial image of the pre-identified user with the image to be identified, a second sub-similarity corresponding to each pre-identified user is obtained; similarly, a third sub-similarity corresponding to each pre-identified user is obtained based on the third sub-database, and a fourth sub-similarity corresponding to each pre-identified user is obtained based on the fourth sub-database;

[0090] (3) For each pre-identified user, a weighted sum is performed based on the first sub-similarity, the second sub-similarity, the third sub-similarity and the fourth sub-similarity to calculate the target similarity between the facial image of each pre-identified user and the facial image of the user to be identified, and then the pre-identified user with the highest target similarity is determined as the target user.

[0091] That is to say, since the pre-identified users are screened according to the preset threshold, it can be considered that there must be a target user among them, so the target similarity of each pre-identified user is calculated, and the pre-identified user with the highest target similarity is determined as the target user; if there are no pre-identified users exceeding the preset threshold in the first sub-database, it can be considered that the user database does not contain the target user, and the user database with other priorities can be replaced to perform face recognition again.

[0092] In this way, by comparing and identifying multiple facial images of the same user, the interference of scene factors on the recognition process can be minimized, thereby increasing the accuracy of facial recognition.

[0093] It should also be noted that if a target user meeting the preset criteria is identified in the user database with a relatively high priority, the identity information of the target user is determined as the identity information of the user to be identified. If the user to be identified cannot be identified in the user database with a relatively high priority, the user database with the next highest priority is searched, and the process continues until all user databases have been traversed.

[0094] It should be noted that face recognition is the process of identifying a target user from multiple user databases according to their priority order. If a stored face image in a user database meets a preset similarity standard with the face image to be identified, the user corresponding to that face image is identified as the target user, and the identity information of the target user is also the identity information of the user to be identified. This completes the face recognition process for the user to be identified.

[0095] In summary, this application proposes a face recognition method, the main contents of which are:

[0096] (1) Based on the historical data of users accessing base stations and the geographical location information of base stations, the probability of users accessing face recognition applications is calculated. When performing face recognition, the search is performed first among the high-probability groups to improve the accuracy of face recognition;

[0097] (2) Based on different types of data accumulated during the operator's business operations, a multi-level face database is established and assigned different weights to perform multiple recognitions to improve the recognition accuracy.

[0098] The main technical effects of this face recognition method are:

[0099] (1) This embodiment not only considers the user's base station access information at the current moment, but also takes into account the base station access information over the past period of time, thereby establishing a relatively static high-probability face information database. This eliminates the need to adjust the face database based on the base station access information.

[0100] (2) This embodiment proposes to establish a hierarchical face database based on images of different quality or sources, and assign different weights to optimize the final face recognition results.

[0101] An embodiment of the present application provides a face recognition method, which obtains an image to be recognized for face recognition using a face recognition application; wherein the image to be recognized includes a face image of a user to be recognized; based on the priorities of multiple user databases, a user database with a relatively high priority among the multiple user databases is preferentially used to perform face recognition on the image to be recognized; wherein the user database stores face images of users; the priority of each user database is determined based on the probability that the user corresponding to the face image stored in each user database will use the face recognition application, and the priority of the user database is positively correlated with the magnitude of the probability. In this way, for a face database with a large data size, it can be divided into multiple user databases according to the probability of users using the face recognition application, and face recognition operations are preferentially performed on user databases with a high usage probability, which can improve the accuracy of the recognition results.

[0102] In another embodiment of the present application, see Figure 3 , which shows a flow chart of the priority determination method provided in an embodiment of the present application, such as Figure 3 As shown, the method may include:

[0103] S201: Acquire facial images of multiple users.

[0104] It should be noted that, by obtaining pre-stored facial images of multiple users, the multiple users are classified and multiple user databases with different priorities are established.

[0105] S202: Calculate the probability that the multiple users use the face recognition application.

[0106] It should be noted that, based on the same concept as the aforementioned facial recognition method, embodiments of the present application provide a priority determination method applied to a user database storing user facial images. Specifically, the application usage probability of multiple users using the facial recognition application is calculated, so that the application usage probability can be used to subsequently form multiple user databases with different priorities.

[0107] The application usage probability is determined based on specified statistical data of a mobile communication network operator providing mobile data services for the face recognition application. Therefore, in some embodiments, the method may further include:

[0108] Determining a target geographic location based on the face recognition application, and obtaining a service area of ​​the face recognition application according to the determined target geographic location; wherein the target geographic location represents the geographic location of the face recognition application when the face recognition application obtains the image to be recognized;

[0109] Determining the multiple base stations according to the service area, and obtaining the specified statistical data based on the multiple base stations; wherein the specified statistical data can represent the size of the probability;

[0110] Determining a target geographic location based on the facial recognition application, and obtaining a service area of ​​the facial recognition application according to the determined target geographic location; wherein the target geographic location represents the geographic location of the facial recognition application when the image to be recognized is obtained;

[0111] Determining the multiple base stations according to the service area, and obtaining the specified statistical data based on the multiple base stations; wherein the specified statistical data can represent the size of the probability;

[0112] The step of calculating the probability that a plurality of users use the face recognition application is performed based on the specified statistical data.

[0113] It should be noted that the process of obtaining specified statistical data can be as follows: first, determine the target geographical location of the face recognition application when it obtains the image to be recognized, and obtain the service area of ​​the face recognition application based on the determined target geographical location; then, determine multiple base stations in the service area, and obtain the aforementioned specified statistical data through multiple base stations.

[0114] Here, the target geographic location refers to the geographic location where the face recognition application is located when it obtains the image to be recognized. In order to establish multiple user databases, a service area is determined based on the radiation radius with the target geographic location as the center, and then all base stations within this service area are determined. For example, the area of ​​500 meters or 1 kilometer around the face recognition device is used as the service area of ​​the face recognition device, and multiple base stations are determined within this service area. Here, the base station refers to the operator base station that can provide data services. In addition, the essence of the base station is to be able to prove that certain people have visited this area, so the base station can also be replaced by other devices that can confirm the user's movement trajectory, such as cash registers, parking fee collection systems, etc. within the service area.

[0115] The specified statistical data includes base station distance data and service probability data. Therefore, in some embodiments, the face recognition application-based acquisition of specified statistical data may include:

[0116] Based on the face recognition application, determining a plurality of base stations that provide mobile data services for the face recognition application;

[0117] Based on the determined multiple base stations, service probability data and base station distance data are obtained; wherein the service probability represents the probability that the multiple base stations provide mobile data services to the user, and the base station distance data represents the distance between the face recognition application and the multiple base stations that can provide mobile data services to the face recognition application.

[0118] It should be noted that the designated statistical data includes base station distance data and service probability data. Therefore, multiple base stations are determined based on the face recognition application, and then the base station distance data and service probability data are obtained. The relevant content of the base station distance data and service probability data is mentioned above and will not be repeated here.

[0119] It should be noted that base station distance data can be easily obtained based on the target geographic location and the deployment location of each base station. However, service probability data requires specific calculations. Therefore, in some embodiments, obtaining service probability data based on the determined multiple base stations may include:

[0120] Acquire the multiple base stations, determine the multiple users, and acquire an access duration set for each user; wherein the access duration set includes a duration for each user to access each base station in the multiple base stations;

[0121] According to the access duration set of each user, the probability that the multiple base stations provide mobile data services for each user is calculated, and the probability that the multiple base stations provide mobile data services for each user is determined as the service probability data.

[0122] It should be noted that the service probability data is not original data, and needs to be calculated based on the historical access information of multiple base stations. Here, the historical access information records the user information that accessed these base stations in a recent historical period, and the access users who accessed the base station in the historical period can be determined based on the historical access information. In addition, the historical access information may include but is not limited to: mobile phone number, International Mobile Equipment Identity (IMEI), Media Access Control Address (MAC) address and access duration. It should be understood that historical access information may refer to access information within the past month, or access information within the past two months or the past six months, and the embodiments of the present application do not make specific limitations.

[0123] It's important to note that, considering access duration and base station distance, if a user spends more time connected to a base station and that base station is closer to the facial recognition device, the probability of that user using the facial recognition device through that base station is higher. Based on this idea, the probability of application usage for a specific user can be calculated based on their access duration and base station distance data.

[0124] Here, base station distance data can be obtained based on multiple base stations and target geographic locations, and service probability data can be calculated based on historical access information of multiple base stations, thereby constituting the aforementioned designated statistical data.

[0125] Specifically, the steps for calculating the application usage probability are as follows: based on the access duration of multiple base stations of each user, determine the probability that the multiple base stations provide mobile data services for each user; based on the probability that the multiple base stations provide mobile data services for each user and the base station distance data, calculate the application usage sub-probability of each user corresponding to each base station; accumulate the application usage sub-probabilities of each user corresponding to each base station to obtain the application usage probability corresponding to each access user.

[0126] In this way, the application usage probability of the user can be calculated based on multiple base stations, so that the access users can be divided into different user databases according to the application usage probability.

[0127] S203: Establishing multiple user databases based on the facial images of the multiple users and the probabilities that the multiple users use the face recognition application.

[0128] It should be noted that multiple users can be grouped based on the probability of application use, thereby establishing multiple user databases containing facial images corresponding to these users. Furthermore, since the priority of each of these multiple user databases is determined based on the probability that the user corresponding to the facial image stored in each database will use the facial recognition application, the priority of each user database is positively correlated with the probability, and the priority of each of the multiple established user databases is also determined.

[0129] Specifically, access users can be classified into several categories based on a preset classification threshold, thereby establishing multiple user databases. The multiple user databases include a high-probability user database and a low-probability user database. Therefore, in some embodiments, the multiple user databases are established based on the facial images of the multiple users and the probability of the multiple users using the face recognition application, and specifically may include:

[0130] Determine a preset classification threshold;

[0131] comparing the probabilities of the plurality of users with the at least one preset classification threshold;

[0132] According to the comparison result, if the probability of one of the users is greater than or equal to the preset classification threshold, the one of the users is classified into the high-probability user database; if the probability of one of the users is less than the preset classification threshold, the one of the users is classified into the low-probability user database.

[0133] It should also be noted that there may be two user databases, namely a high probability user database and a low probability user database. In this case, by comparing the user's probability with a preset classification threshold, the user can be divided into the high probability user database and the low probability user database respectively.

[0134] Of course, the number of user databases can be three, namely a high probability user database, a medium probability user database and a low probability user database. In this case, a first preset classification threshold and a second preset classification threshold are determined respectively, and the first preset classification threshold is higher than the second preset classification threshold.

[0135] Then, the probability of each user is compared with the first preset classification threshold and the second preset classification threshold; based on the comparison result, users whose probability is greater than or equal to the first preset classification threshold are classified into the high-probability user database; users whose probability is greater than or equal to the second preset classification threshold and less than the first preset classification threshold are classified into the medium-probability user database; users whose probability is less than the second preset classification threshold are classified into the low-probability user database.

[0136] In addition, the number of users in the user database can be greater, and this embodiment of the present application does not specifically limit it.

[0137] It should also be noted that when establishing multiple user databases, facial feature vectors can be calculated separately for each user database, so that the specific parameters of the convolutional neural network used for face recognition operations in different user databases are different. Alternatively, facial feature vectors can be calculated uniformly for images in all user databases, so that face recognition operations can be performed using convolutional neural networks with the same parameters in different user databases.

[0138] Through the above steps, multiple user databases can be established, and the probability of user groups in each user database accessing the face recognition application is different, so that the image to be recognized can be recognized first in the high-probability user database to improve the recognition accuracy.

[0139] An embodiment of the present application provides a priority determination method, which comprises obtaining facial images of multiple users; calculating the probability of the multiple users using a facial recognition application; and establishing multiple user databases based on the facial images of the multiple users and the probability of the multiple users using the facial recognition application. The priority of each of the multiple user databases is determined based on the probability of the user corresponding to the facial image stored in each user database using the facial recognition application, and the priority of the user database is positively correlated with the magnitude of the probability. In this way, for a facial database with a large data size, it is possible to divide it into multiple user databases based on the probability of users using the facial recognition application, and prioritize facial recognition operations in user databases with a high usage probability, thereby improving the accuracy of the recognition results.

[0140] In another embodiment of the present application, see Figure 5 , which shows a flow chart of another face recognition method provided by an embodiment of the present application, such as Figure 5 As shown, the method may include:

[0141] S301: According to the service scope of the face recognition application, determine multiple base stations that can provide data services and record them as b1, b2...b n , respectively count the distances d1, d2…d from the multiple base stations to the face recognition application deployment location n ;

[0142] It should be noted that, first, the service range of the face recognition application is determined with the location where the target face recognition application is deployed as the center, and multiple base stations that may provide mobile data services are determined within the determined range, and these base stations are respectively marked as b1, b2...b n There are n base stations in total, and the distances between these n base stations and the deployment location of the face recognition application are d1, d2…d n .

[0143] S302: Calculate the probability p(i) of a user accessing base station i based on the historical access data of the past month;

[0144] It should be noted that the user information accessed to the above n base stations in the past month is collected, including but not limited to mobile phone number, IMEI, MAC address and access duration. The connection duration of the user to the above n base stations is t1…t n , the total time is t, then the probability p(i) of user accessing base station i can be calculated by the following formula:

[0145] S303: Calculate the probability y(i) that the user accesses the face recognition application when accessing base station i, and calculate the probability y that the user accesses the face recognition application through n base stations;

[0146] It should be noted that the probability of a user accessing a face recognition application is related to the user's access to a specific base station and the distance to the base station. If the user has a high probability of accessing a base station and the base station is close to the application deployment site, then the user is likely to access the deployed application. If the user has a low probability of accessing a base station and the base station is far from the application deployment site, then the user's probability of accessing the application will also be low. Let y(i) represent the probability of a user accessing an application when accessing base station bi. The specific calculation formula can be: Further calculations are performed to obtain the probability of a user actually accessing an application when accessing through n base stations:

[0147] S304: Establishing multiple different face databases based on the probability of users accessing the face recognition application.

[0148] It should be noted that multiple different face databases are established for all users within the service scope of the face recognition application based on the probability of accessing the application. When performing face comparison, priority is given to searching in the high-probability face database. If no face is found, the search is then performed in the low-probability face database.

[0149] The embodiment of the present application provides a face recognition method. Through this embodiment, the specific implementation of the aforementioned embodiment is elaborated in detail. It can be seen that for a face database with a large data scale, it can be divided into multiple user databases according to the probability of users using face recognition applications. Face recognition operations are performed preferentially in user databases with a high usage probability, which can improve the accuracy of the recognition results.

[0150] In yet another embodiment of the present application, see Figure 6 , which shows a flow chart of another face recognition method provided by an embodiment of the present application, such as Figure 6 As shown, the method may include:

[0151] S401: establishing a first face base library, a second face base library, a third face base library, and a fourth face base library respectively according to images from different sources;

[0152] It should be noted that according to the aforementioned embodiment, multiple face databases containing different user groups are established based on the probability of users accessing applications. For each specific face database, multiple face base databases with different image sources are included.

[0153] It should also be noted that the multiple face databases may include four categories, namely the first face database, the second face database, the third face database, and the fourth face database. The users included in the four face databases are the same, but the images of the same user in each face database come from different sources. Specifically, the user images in the first face database may be images retained by users when they conduct business on-site; the user images in the second face database may be images uploaded by users when they conduct business remotely; the user images in the third face database may be images of users captured by a camera; and the user images in the fourth face database may be images of ID documents uploaded by users when they conduct business or images of ID documents captured using a scanner or other equipment.

[0154] That is to say, in face recognition applications, a face database is usually pre-established based on the actual application situation, and face recognition is performed based on this face database. In the operator application scenario, the construction of the face database can use the existing data (Note: this part of the data involves user privacy and requires strict confidentiality policies to prevent data leakage) to initialize the face database. When establishing the face feature database, multiple different levels of face databases are established based on images from different sources, including: (1) images retained by users when handling business on-site; (2) images uploaded by users who handle business remotely; (3) user images captured by cameras; (4) ID images uploaded by users when handling business or ID images collected using devices such as scanners.

[0155] S402: Searching the first face recognition database using the image to be recognized. If a first sub-similarity between the image to be recognized and one of the user images is greater than a first preset threshold, retaining the first sub-similarity and recording it as s1 corresponding to the one of the user images.

[0156] It should be noted that when searching the face database of the first category of images, images with a similarity below the threshold t are directly excluded, while images with a similarity above t are retained and recorded as s1. When the threshold t is low or the image clarity is low, multiple user images with similarities above the threshold t may be retrieved. For each of these user images with similarities above t, s1 is recorded and these user images from other sources are compared with the face image to be identified.

[0157] S403: For a user whose first sub-similarity is greater than a first preset threshold, compare the user's image in the second face database, the third face database, and the fourth face database with the face image to be identified, and record the similarities as s2, s3, and s4 respectively;

[0158] It should be noted that when searching in the face database of the second, third and fourth categories of images, the faces excluded in the first step can be directly skipped, and the similarities obtained for each category of images are recorded as s2, s3, and s4 respectively.

[0159] S404: Calculate the comprehensive similarity: s = s1*0.4+s2*0.3+s3*0.3+s4*0.1.

[0160] It should be noted that since the clarity and reliability of user images in each face database are different, each type of image has a different weight when calculating the comprehensive similarity, and the final sum is used to obtain the overall similarity. Finally, the user with the highest overall similarity is taken as the final recognition result.

[0161] The embodiments of the present application provide a face recognition method. The specific implementation of the aforementioned embodiment is described in detail through the above embodiments. It can be seen that for a face database with a large data scale, it can be divided into multiple user databases according to the probability of users using face recognition applications. Face recognition operations are performed preferentially in user databases with a high usage probability, which can improve the accuracy of the recognition results.

[0162] In another embodiment of the present application, see Figure 7 , which shows a schematic diagram of the composition structure of a face recognition device 50 provided in an embodiment of the present application, such as Figure 7 As shown, the face recognition device 50 includes an acquisition unit 501, a recognition unit 502 and a result unit 503, wherein:

[0163] An acquisition unit 501 is configured to acquire an image to be recognized for performing a face recognition operation using a face recognition application; wherein the image to be recognized includes a face image of a user to be recognized;

[0164] The identification unit 502 is configured to perform a face recognition operation on the image to be identified based on the priorities of multiple user databases, preferentially utilizing a user database with a relatively higher priority among the multiple user databases; wherein the user database stores facial images of users; the priority of each user database is determined according to the probability that the user corresponding to the facial image stored in each user database uses the face recognition application, and the priority of the user database is positively correlated with the size of the probability.

[0165] In some embodiments, the probability is determined based on specified statistical data of a mobile communication network operator that provides mobile data services for the face recognition application; wherein the specified statistical data can characterize the size of the probability.

[0166] In some embodiments, the specified statistical data includes: base station distance data and service probability data; wherein, the base station distance data represents the distance between the face recognition application and multiple base stations that can provide mobile data services for the face recognition application; the service probability data represents the probability that the multiple base stations provide mobile data services to the user.

[0167] In some embodiments, the facial images of users stored in the user database include: a high-quality facial image set consisting of high-quality facial images, and a low-quality facial image set consisting of low-quality facial images; the high-quality facial images refer to facial images retained when users handle services that have relatively high requirements for facial image quality; the low-quality facial images refer to facial images retained when users handle services that have relatively low requirements for facial image quality.

[0168] In some embodiments, the recognition unit 502 is configured to determine a user database that is preferentially used for performing a face recognition operation on the image to be recognized; based on the determined high-quality face image set and low-quality face image set in the user database, perform face recognition on the image to be recognized, respectively, to obtain a first face similarity judgment result corresponding to the high-quality face image set and a second face similarity judgment result corresponding to the low-quality face image set;

[0169] The face recognition result of the image to be recognized is obtained by fusing the first face similarity judgment result and the second face similarity judgment result.

[0170] It is understood that in this embodiment, a "unit" may be a portion of a circuit, a portion of the first processor, a portion of a program or software, and so forth. It may also be a module or a non-modular system. Furthermore, the various components in this embodiment may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional modules.

[0171] If the integrated unit is implemented as a software functional module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment, or the portion that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the method described in this embodiment. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0172] Therefore, this embodiment provides a computer storage medium, which stores a face recognition program. When the face recognition program is executed by at least one processor, the steps of the method described in any one of the above embodiments are implemented.

[0173] Based on the composition of the above-mentioned face recognition device 50 and the computer storage medium, see Figure 8 , which shows a specific hardware structure example of a face recognition device 50 provided in an embodiment of the present application, such as Figure 8 As shown, the face recognition device 50 may include: a first communication interface 601, a first memory 602 and a first processor 603; each component is coupled together via a bus device 604. It is understood that the bus device 604 is used to achieve connection and communication between these components. In addition to the data bus, the bus device 604 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, Figure 8 Various buses are labeled as bus devices 604. Among them, the first communication interface 601 is used to receive and send signals during the process of sending and receiving information between other external network elements;

[0174] A first memory 602 is used to store computer programs that can be run on the first processor 603;

[0175] The first processor 603 is configured to, when running the computer program, execute:

[0176] Acquire an image to be recognized for performing a face recognition operation using a face recognition application; wherein the image to be recognized includes a face image of a user to be recognized;

[0177] Based on the priorities of multiple user databases, a user database with a relatively higher priority among the multiple user databases is preferentially used to perform a face recognition operation on the image to be recognized; wherein, the user database stores the face images of users; the priority of each user database is determined according to the probability that the user corresponding to the face image stored in each user database uses the face recognition application, and the priority of the user database is positively correlated with the size of the probability.

[0178] It is understood that the first memory 602 in the embodiment of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus random access memory (DRRAM). The first memory 602 of the apparatus and method described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0179] The first processor 603 may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by hardware integrated logic circuits or software instructions in the first processor 603. The above-mentioned first processor 603 can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The various methods, steps, and logic block diagrams disclosed in the embodiments of this application can be implemented or executed. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this application can be directly implemented and executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium mature in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in the first memory 602 , and the first processor 603 reads the information in the first memory 602 and completes the steps of the above method in combination with its hardware.

[0180] It is understood that the embodiments described herein may be implemented using hardware, software, firmware, middleware, microcode, or a combination thereof. For hardware implementation, the processing unit may be implemented in one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers, microprocessors, or other electronic units or combinations thereof for performing the functions described herein.

[0181] For software implementation, the techniques described herein can be implemented by modules (e.g., procedures, functions, etc.) that perform the functions described herein. The software code can be stored in a memory and executed by a processor. The memory can be implemented in the processor or external to the processor.

[0182] Optionally, as another embodiment, the first processor 603 is further configured to execute the steps of the method in any one of the aforementioned embodiments when running the computer program.

[0183] In another embodiment of the present application, see Figure 9 , which shows a structural diagram of a priority determination device 70 provided in an embodiment of the present application, such as Figure 9 As shown, the priority determination device 70 includes an image unit 701, a calculation unit 702 and a determination unit 703; wherein,

[0184] An imaging unit 701 is configured to obtain facial images of multiple users;

[0185] a calculation unit 702 configured to calculate the probability of multiple users using a face recognition application;

[0186] The determination unit 703 is configured to establish multiple user databases based on the facial images of the multiple users and the probability of the multiple users using the facial recognition application; wherein the priority of each of the multiple user databases is determined according to the probability of the users corresponding to the facial images stored in each user database using the facial recognition application, and the priority of the user database is positively correlated with the size of the probability.

[0187] In some embodiments, the probability is determined based on specified statistical data of a mobile communication network operator that provides mobile data services for the face recognition application; the computing unit 702 is further configured to determine a target geographic location based on the face recognition application, and obtain a service area of ​​the face recognition application based on the determined target geographic location; wherein the target geographic location represents the geographic location of the face recognition application when the image to be recognized is obtained; the multiple base stations are determined based on the service area, and the specified statistical data are obtained based on the multiple base stations; wherein the specified statistical data can characterize the size of the probability; based on the specified statistical data, the step of calculating the probability of multiple users using the face recognition application is performed.

[0188] In some embodiments, the specified statistical data includes base station distance data and service probability data; the computing unit 702 is further configured to determine, based on the face recognition application, multiple base stations that provide mobile data services for the face recognition application; and obtain service probability data and base station distance data based on the determined multiple base stations; wherein the service probability represents the probability that the multiple base stations provide mobile data services to the user, and the base station distance data represents the distance between the face recognition application and multiple base stations that can provide mobile data services for the face recognition application.

[0189] In some embodiments, the calculation unit 702 is further configured to determine the multiple users based on the multiple base stations, and obtain an access duration set for each user; wherein the access duration set includes the duration for each user to access each base station in the multiple base stations; based on the access duration set for each user, calculate the probability that the multiple base stations provide mobile data services for each user, and determine the probability that the multiple base stations provide mobile data services for each user as the service probability data.

[0190] In some embodiments, the multiple user databases include a high-probability user database and a low-probability user database; the determination unit 703 is further configured to determine a preset classification threshold; compare the probabilities of the multiple users with the at least one preset classification threshold; based on the comparison result, if the probability of one of the users is greater than or equal to the preset classification threshold, classify the one of the users into the high-probability user database; if the probability of one of the users is less than the preset classification threshold, classify the one of the users into the low-probability user database.

[0191] It is understood that in this embodiment, a "unit" can be a portion of a circuit, a portion of a processor, a portion of a program or software, etc., and can also be a module or a non-modular system. Furthermore, the various components in this embodiment can be integrated into a single processing unit, or each unit can exist physically separately, or two or more units can be integrated into a single unit. The aforementioned integrated units can be implemented in the form of hardware or software functional modules.

[0192] If the integrated unit is implemented as a software functional module and not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, this embodiment provides a computer storage medium that stores a priority determination program. When the priority determination program is executed by the second processor, it implements any of the methods in the aforementioned embodiments.

[0193] Based on the above-mentioned composition of the priority determination device 70 and the computer storage medium, see Figure 10 , which shows a specific hardware structure example of the priority determination device 70 provided in an embodiment of the present application, such as Figure 10 As shown, the proxy node 50 may include: a second communication interface 801, a second memory 802, and a second processor 803; each component is coupled together via a second bus system 804. It is understood that the second bus system 804 is used to implement connection and communication between these components. In addition to the data bus, the second bus system 804 also includes a power bus, a control bus, and a status signal bus. However, for the sake of clarity, Figure 10 In FIG, various buses are labeled as a second bus system 804.

[0194] The second communication interface 801 is used to receive and send signals during the process of sending and receiving information with other external network elements;

[0195] The second memory 802 is used to store computer programs that can be run on the second processor 803;

[0196] The second processor 803 is configured to, when running the computer program, execute:

[0197] Calculate the probability of multiple users using face recognition applications;

[0198] Based on the probabilities and the facial images corresponding to the multiple users, multiple user databases are established, and the priorities of the multiple user databases are determined based on the probabilities of the users corresponding to the facial images stored in each of the multiple user databases; wherein the priorities of the user databases are positively correlated with the magnitude of the probabilities.

[0199] Optionally, as another embodiment, the second processor 803 is further configured to execute any one of the methods described in the foregoing embodiments when running the computer program.

[0200] It can be understood that the hardware functions of the second memory 802 and the first memory 602 are similar, and the hardware functions of the second processor 803 and the first processor 603 are similar; they will not be described in detail here.

[0201] Based on the composition and hardware structure diagram of the above-mentioned face recognition device 50 and priority determination device 70, see Figure 11 , which shows a schematic diagram of the composition structure of a face recognition system 90 provided in an embodiment of the present application. Figure 11 As shown, the face recognition system 90 includes at least the face recognition device 50 and the priority determination device 70 according to any one of the aforementioned embodiments.

[0202] For the face recognition system 90, for a face database with a large data size, it can be divided into multiple user databases according to the usage probability of the face recognition device, and face recognition is performed preferentially in the user database with a high usage probability, which can improve the accuracy of the recognition results.

[0203] The above description is merely a preferred embodiment of the present application and is not intended to limit the scope of protection of the present application.

[0204] It should be noted that, in this application, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0205] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0206] The methods disclosed in the several method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0207] The features disclosed in the several product embodiments provided in this application can be arbitrarily combined without conflict to obtain new product embodiments.

[0208] The features disclosed in the several method or device embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0209] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A face recognition method, characterized in that: The method comprises: Acquire an image to be recognized for performing a face recognition operation using a face recognition application; wherein the image to be recognized includes a face image of a user to be recognized; Based on the priorities of multiple user databases, preferentially utilizing a user database with a relatively higher priority among the multiple user databases to perform a face recognition operation on the image to be recognized; wherein the user database stores facial images of users; the priority of each user database is determined based on the probability that the user corresponding to the facial image stored in the user database uses the face recognition application, and the priority of the user database is positively correlated with the magnitude of the probability; The probability of the user using the face recognition application is determined based on the base station distance data and service probability data of multiple base stations around the face recognition application. The service probability data of the base station for a user refers to the ratio of the user's access time to the base station in the historical access data to the total access time of the user to the multiple base stations.

2. The face recognition method according to claim 1, characterized in that The probability is determined based on specified statistical data of a mobile communication network operator providing mobile data services for the face recognition application; The specified statistical data can represent the size of the probability.

3. The face recognition method according to claim 2, characterized in that The specified statistical data includes: base station distance data and service probability data; wherein, The base station distance data represents the distance between the face recognition application and a base station capable of providing mobile data services for the face recognition application; The service probability data indicates the probability that the base station provides the mobile data service for the user.

4. The face recognition method according to claim 1, characterized in that The facial images of users stored in the user database include: a high-quality facial image set consisting of high-quality facial images, and a low-quality facial image set consisting of low-quality facial images; The high-quality facial image refers to the facial image retained when the user handles a business with relatively high requirements for facial image quality; the low-quality facial image refers to the facial image retained when the user handles a business with relatively low requirements for facial image quality.

5. The face recognition method according to claim 4, characterized in that: The step of preferentially utilizing a user database having a relatively higher priority among the multiple user databases to perform a face recognition operation on the image to be recognized includes: Determining a user database that is preferentially used for performing a face recognition operation on the image to be recognized; Based on the determined set of high-quality facial images and the set of low-quality facial images in the user database, performing facial recognition on the images to be recognized, respectively, to obtain a first facial similarity judgment result corresponding to the set of high-quality facial images and a second facial similarity judgment result corresponding to the set of low-quality facial images; The face recognition result of the image to be recognized is obtained by fusing the first face similarity judgment result and the second face similarity judgment result.

6. A priority determination method, characterized in that: The method comprises: Obtain facial images of multiple users; Calculating probabilities that the multiple users use the face recognition application; Based on the facial images of the multiple users and the probabilities of the multiple users using the facial recognition application, multiple user databases are established; wherein the priority of each of the multiple user databases is determined according to the probability of the user corresponding to the facial image stored in each user database using the facial recognition application; the priority of the user database is positively correlated with the size of the probability; the probability of the user using the facial recognition application is determined based on the base station distance data and service probability data of multiple base stations around the facial recognition application, and the service probability data of the base station for a user is the ratio of the user's access time to the base station to the total access time of the base station.

7. The priority determination method according to claim 6, characterized in that: The probability is determined based on specified statistical data of a mobile communication network operator providing mobile data services for the face recognition application; the method further comprising: Determining a target geographic location based on the facial recognition application, and obtaining a service area of ​​the facial recognition application according to the determined target geographic location; wherein the target geographic location represents the geographic location of the facial recognition application when the image to be recognized is obtained; Determining the multiple base stations according to the service area, and obtaining the specified statistical data based on the multiple base stations; wherein the specified statistical data can represent the size of the probability; The step of calculating the probability that a plurality of users use the face recognition application is performed based on the specified statistical data.

8. The priority determination method according to claim 7, characterized in that: The specified statistical data includes base station distance data and service probability data; and obtaining the specified statistical data based on the multiple base stations includes: Based on the face recognition application, determining a plurality of base stations that provide mobile data services for the face recognition application; Based on the determined multiple base stations, service probability data and base station distance data are obtained; wherein the service probability represents the probability that the multiple base stations provide mobile data services to the user, and the base station distance data represents the distance between the face recognition application and the multiple base stations that can provide mobile data services to the face recognition application.

9. The priority determination method according to claim 8, characterized in that: The acquiring service probability data according to the determined multiple base stations includes: Determine the multiple users according to the multiple base stations, and obtain an access duration set for each user; wherein the access duration set includes a duration for each user to access each of the multiple base stations; According to the access duration set of each user, the probability that the multiple base stations provide mobile data services for each user is calculated, and the probability that the multiple base stations provide mobile data services for each user is determined as the service probability data.

10. The priority determination method according to claim 6, wherein: The multiple user databases include a high-probability user database and a low-probability user database; the multiple user databases are established based on the facial images of the multiple users and the probabilities of the multiple users using the face recognition application, including: Determine a preset classification threshold; comparing the probabilities of the plurality of users with the at least one preset classification threshold; According to the comparison result, if the probability of one of the users is greater than or equal to the preset classification threshold, the one of the users is classified into the high-probability user database; if the probability of one of the users is less than the preset classification threshold, the one of the users is classified into the low-probability user database.

11. A face recognition device, characterized in that: The face recognition device includes a receiving unit and a recognition unit, wherein: The receiving unit is configured to obtain an image to be recognized for performing a face recognition operation using a face recognition application; wherein the image to be recognized includes a face image of a user to be recognized; The identification unit is configured to perform face recognition operations on the image to be identified based on the priorities of multiple user databases, giving priority to using user databases with relatively higher priorities among the multiple user databases; wherein the user databases store facial images of users; the priorities of each user database are determined according to the probability that the user corresponding to the facial image stored in each user database uses the face recognition application, and the priority of the user database is positively correlated with the size of the probability; the probability of the user using the face recognition application is determined based on the base station distance data and service probability data of multiple base stations around the face recognition application, and the service probability data of the base station for a user refers to the ratio of the access time of the user to the base station in the historical access data to the total access time of the user to the multiple base stations.

12. A face recognition device, characterized in that: The face recognition device includes a first memory and a first processor; wherein, The first memory is used to store a computer program that can be run on the first processor; The first processor is configured to execute the steps of the method according to any one of claims 1 to 5 when running the computer program.

13. A priority determination device, characterized in that: The priority determination device includes an image unit, a calculation unit and a determination unit, wherein: The imaging unit is configured to obtain facial images of multiple users; The calculation unit is configured to calculate the probability of multiple users using the face recognition application; The determination unit is configured to establish multiple user databases based on the facial images of the multiple users and the probability of the multiple users using the facial recognition application; wherein the priority of each of the multiple user databases is determined according to the probability of the user corresponding to the facial image stored in each user database using the facial recognition application, and the priority of the user database is positively correlated with the size of the probability; the probability of the user using the facial recognition application is determined based on the base station distance data and service probability data of multiple base stations around the facial recognition application, and the service probability data of the base station for a user is obtained by using the ratio of the user's access time to the base station to the total access time of the base station.

14. A priority determination device, characterized in that: The priority determination device includes a second memory and a second processor; wherein, The second memory is used to store a computer program that can be run on the second processor; The second processor is configured to execute the steps of the method according to any one of claims 6 to 11 when running the computer program.

15. A computer storage medium, characterized in that The computer storage medium stores a face recognition program or a priority determination program; wherein, When the face recognition program is executed by at least one processor, the steps of the method according to any one of claims 1 to 5 are implemented; When the priority determination program is executed by at least one processor, the steps of the method according to any one of claims 6 to 10 are implemented.

16. A face recognition system, characterized in that: The face recognition system at least includes the face recognition device according to claim 11 or 12, and the priority determination device according to claim 13 or 14.

Images