WO2022166207A1

WO2022166207A1 - Face recognition method and apparatus, device, and storage medium

Info

Publication number: WO2022166207A1
Application number: PCT/CN2021/118143
Authority: WO
Inventors: 康家梁; 卞凯; 傅宜生; 冀乃庚
Original assignee: 中国银联股份有限公司
Priority date: 2021-02-07
Filing date: 2021-09-14
Publication date: 2022-08-11
Also published as: TWI789128B; CN112818885A; TW202232367A

Abstract

The present application discloses a face recognition method and apparatus, a device, and a storage medium. The face recognition method comprises: obtaining a first face image of a target recognition object; determining a face matching degree between the target recognition object and each predetermined object according to the first face image and a face image of each predetermined object in a first face library; when the face matching degree satisfies a first predetermined condition, obtaining first M face matching degrees from a face matching degree sequence, wherein the first predetermined condition comprises at least one of the following: the face matching degrees between the target recognition object and the plurality of predetermined objects are greater than a first predetermined threshold, and a difference between the maximum face matching degree and the first predetermined threshold is less than a predetermined value; calculating a difference between every two adjacent face matching degrees in the M face matching degrees to obtain M-1 differences; and according to the M-1 differences, determining, in the plurality of predetermined objects, a first object matched with the face of the target recognition object. The face recognition method disclosed in the present application can reduce the risk of misrecognition.

Description

Face recognition method, device, equipment and storage medium

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the priority of Chinese Patent Application No. 202110175970.5 filed on February 07, 2021, entitled "Facial Recognition Method, Apparatus, Equipment and Storage Medium", the entire content of which is incorporated herein by reference.

technical field

The present application relates to the field of computer technology, and in particular, to a face recognition method, apparatus, device, and storage medium.

Background technique

Face recognition is a hot research and application direction in the field of artificial intelligence vision. This technology has been widely used in commercial passenger flow analysis, security monitoring, mobile phone applications, and institutional information verification and comparison.

The face recognition scheme in the related art is to calculate the similarity between the face image collected on the spot and each face image in the face database, and the face image collected on the spot is similar to a certain face image in the face database. If the degree is greater than the predetermined threshold, the two are considered to match, otherwise, the two are considered to be unmatched.

However, the above face recognition scheme is relatively simple, and the risk of misrecognition is relatively high.

SUMMARY OF THE INVENTION

Embodiments of the present application provide a face recognition method, apparatus, device, and storage medium, which can solve the technical problem of a relatively high risk of misrecognition during face recognition.

In a first aspect, an embodiment of the present application provides a face recognition method, including:

Obtain the first face image of the target recognition object;

Determine, according to the first face image and the face images of each predetermined object in the first face database, the degree of face matching between the target recognition object and each of the predetermined objects;

In the case that the face matching degree satisfies the first predetermined condition, the first M face matching degrees are obtained from the face matching degree sequence, and the face matching degree sequence is the target recognition object and each of the predetermined objects. M is an integer greater than 1, and the first predetermined condition includes at least one of the following: the target recognition object and a plurality of people of the predetermined object The face matching degree is greater than a first predetermined threshold, and the difference between the maximum face matching degree and the first predetermined threshold is less than a predetermined value;

Calculate the difference value of each adjacent two described human face matching degrees in the M described human face matching degrees, and obtain M-1 difference values;

According to the M-1 difference values, a first object matching the face of the target recognition object is determined from among the plurality of predetermined objects.

In a second aspect, an embodiment of the present application provides a face recognition device, including:

a first acquisition module, used for acquiring the first face image of the target recognition object;

a first determination module, configured to determine the face matching degree between the target recognition object and each of the predetermined objects according to the first face image and the face images of each predetermined object in the first face database;

The second obtaining module is configured to obtain the first M face matching degrees from the face matching degree sequence when the face matching degree satisfies the first predetermined condition, and the face matching degree sequence is the target recognition A sequence obtained by arranging the face matching degrees of the objects and each of the predetermined objects in descending order; M is an integer greater than 1, and the first predetermined condition includes at least one of the following: the target recognition object and multiple The face matching degree of each of the predetermined objects is greater than a first predetermined threshold, and the difference between the largest face matching degree and the first predetermined threshold is less than a predetermined value;

The first calculation module is used to calculate the difference between each adjacent two described face matching degrees in the M described face matching degrees, and obtain M-1 difference values;

The second determination module is configured to determine, according to the M−1 difference values, a first object matching the face of the target recognition object among the plurality of predetermined objects.

In a third aspect, an embodiment of the present application provides a face recognition device, the device includes: a processor and a memory storing computer program instructions, the processor implements the first aspect when the computer program instructions are executed. face recognition method.

In a fourth aspect, an embodiment of the present application provides a computer storage medium, where computer program instructions are stored thereon, and when the computer program instructions are executed by a processor, the face recognition method of the first aspect is implemented.

According to the face recognition method, device, device, and storage medium of the embodiments of the present application, after determining the face matching degree between the target recognition object and each predetermined object in the face database, if the target recognition object and multiple predetermined objects appear If the face matching degree is greater than the first predetermined threshold, it means that there are multiple predetermined objects and the face matching degree of the target recognition object is relatively high. Also, if the difference between the maximum face matching degree and the first predetermined threshold is smaller than the predetermined value, it means that the maximum face matching degree is relatively close to the first predetermined threshold. In the above case, there is a risk of misidentification. Therefore, a larger M face matching degree is obtained, and then, the difference between the matching degrees of each adjacent two faces in the M face matching degree is calculated, and M-1 difference values are obtained; according to the M-1 difference values, A first object matching the face of the target recognition object is determined among the plurality of predetermined objects. Since the face recognition is not only performed by judging whether the maximum face matching degree is greater than the threshold, but more factors are considered in the face recognition process, so that the results of face recognition are more accurate and the risk of misrecognition is reduced.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present application more clearly, the following briefly introduces the drawings required in the embodiments of the present application. Obviously, the drawings described below are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can also be obtained from these drawings without any creative effort.

FIG. 1 is a schematic flowchart of an embodiment of a face recognition method provided by the present application.

FIG. 2 is a schematic flowchart of another embodiment of a face recognition method provided by the present application.

FIG. 3 is a schematic flowchart of another embodiment of a face recognition method provided by the present application.

FIG. 4 is a schematic flowchart of still another embodiment of a face recognition method provided by the present application.

FIG. 5 is a schematic structural diagram of an embodiment of a face recognition device provided by the present application.

FIG. 6 shows a schematic diagram of a hardware structure of an embodiment of a face recognition device provided by the present application.

Detailed ways

The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application more clear, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, but not to limit the present application. It will be apparent to those skilled in the art that the present application may be practiced without some of these specific details. The following description of the embodiments is merely to provide a better understanding of the present application by illustrating examples of the present application.

It should be noted that, in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any relationship between these entities or operations. any such actual relationship or sequence exists. Moreover, the terms "comprising", "comprising" or any other variation thereof are intended to encompass a non-exclusive inclusion such that a process, method, article or device that includes a list of elements includes not only those elements, but also includes not explicitly listed or other elements inherent to such a process, method, article or apparatus. Without further limitation, an element defined by the phrase "comprising" does not preclude the presence of additional identical elements in the process, method, article or device that includes the element.

With the development of face recognition technology, face recognition has been widely used in various fields. In the face recognition process, the face recognition request is first sent to the biometrics-related platform through the face routing gateway, and then the biometrics-related platform performs face recognition. In the face recognition process, efficient and accurate recognition services are becoming more and more important.

In the related art, the face image of each predetermined object is pre-stored in the face database. When performing face recognition on the target recognition object, the face image of the target recognition object is first collected, and then, according to the collected face image and the human face The face images of each predetermined object in the face database are used to calculate the face matching degree between the target recognition object and each predetermined object. If there is a face matching degree greater than a certain threshold, the two are considered to match, otherwise, the two are considered not to match.

Different recognition strategies may perform different optimizations based on the above process. For example, when performing face retrieval in the face database, if there is only one predetermined face in the face database and the matching degree of the face of the target recognition object exceeds the threshold, it will be considered as not. Risks occur as a result of unambiguous identification matches.

However, in the case of a large number of face images in the face database, the risk of misrecognition is prone to occur, that is, there may be multiple face matching degrees that are all greater than the threshold, or the matching degrees of the faces that are greater than the threshold are relatively close to the threshold. In these cases, misrecognition is prone to occur, leading to a high risk of misidentification of faces.

In order to solve the technical problem that the risk of face misrecognition is relatively high, the present application provides a face recognition method, which can be applied to a face recognition device. The face recognition device can be a mobile electronic device or a non-mobile electronic device. Exemplarily, the mobile electronic device may be a mobile phone, a tablet computer, a notebook computer, a palmtop computer, an in-vehicle electronic device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (personal digital assistant). assistant, PDA), etc., non-mobile electronic devices can be servers, network attached storage (Network Attached Storage, NAS), personal computer (personal computer, PC), television (television, TV), teller machine or self-service machine, etc., this application Examples are not specifically limited.

In the face recognition method, first obtain the first face image of the target recognition object; according to the first face image and the face images of each predetermined object in the first face database, determine the target recognition object and the person of each predetermined object face matching.

After determining the face matching degree between the target recognition object and each predetermined object in the first face database, if the face matching degree between the target recognition object and multiple predetermined objects is greater than the first predetermined threshold, it means that there are multiple predetermined The face matching degree of the object and the target recognition object is relatively high. Also, if the difference between the maximum face matching degree and the first predetermined threshold is smaller than the predetermined value, it means that the maximum face matching degree is relatively close to the first predetermined threshold. For the above cases, there is a risk of misidentification.

Therefore, a larger M face matching degree is obtained, and then, the difference between the matching degrees of each adjacent two faces in the M face matching degree is calculated, and M-1 difference values are obtained; according to the M-1 difference values, A first object matching the face of the target recognition object is determined among the plurality of predetermined objects. Since it is not only to judge whether the maximum face matching degree is greater than the threshold for face recognition, but more factors are considered in the face recognition process, so that the result of face recognition can be made more accurate and the number of misidentifications can be reduced. risk.

The following describes the face recognition method provided by the present application with reference to the accompanying drawings. FIG. 1 is a schematic flowchart of an embodiment of a face recognition method provided by the present application.

As shown in FIG. 1, the face recognition method 100 includes:

S102: Acquire a first face image of the target recognition object.

In S102, the first face image sent from the electronic device may be received. For example, when a user makes payment using an electronic device, the electronic device collects the user's face, obtains a first face image, and sends the first face image to the face recognition device. The face recognition device receives the first face image sent by the electronic device.

The face recognition method 100 further includes: S104, according to the first face image and the face images of each predetermined object in the first face database, determine the face matching degree between the target recognition object and each predetermined object.

In S104, the similarity between the first face image and the face images of each predetermined object in the first face database may be calculated, and the face matching degree between the target recognition object and each predetermined object is determined according to the similarity.

The face recognition method 100 further includes: S106, in the case that the face matching degree satisfies the first predetermined condition, obtain the first M face matching degree from the face matching degree sequence, and the face matching degree sequence is the target recognition object and each face matching degree. A sequence obtained by arranging the face matching degrees of the predetermined objects in descending order; M is an integer greater than 1, and the first predetermined condition includes at least one of the following: the face matching degrees of the target recognition object and multiple predetermined objects are greater than For the first predetermined threshold, the difference between the maximum face matching degree and the first predetermined threshold is less than a predetermined value.

In S106, in the case that the face matching degree between the target recognition object and multiple predetermined objects is greater than the first predetermined threshold, it means that there are multiple predetermined objects and the face matching degree of the target recognition object is relatively high, that is to say, there are many The face of the predetermined object is similar to the face of the target recognition object. In this case, the risk of misidentification is more likely to occur.

Also, if the difference between the maximum face matching degree and the first predetermined threshold is smaller than the predetermined value, it means that the difference between the maximum face matching degree and the first predetermined threshold is relatively small, for example, the maximum face matching degree is in the first predetermined threshold. The threshold fluctuates within a range of 0.5 above and below the threshold. This situation may be due to the influence of the environment and the angle of collecting face images, resulting in that the matching degree of the target recognition object and each predetermined object is not very high. In this case, the risk of misrecognition is also prone to occur.

In view of the above situation, first, the calculated face matching degrees are in descending order to obtain a face matching degree sequence, and then, the first M face matching degrees are obtained from the face matching degree sequence.

The face recognition method 100 further includes: S108, calculating the difference between the matching degrees of each adjacent two faces in the matching degrees of M faces, and obtaining M-1 difference values;

S110, according to the M-1 difference values, determine a first object matching the face of the target recognition object among the plurality of predetermined objects.

In the embodiment of the present application, since it is not only judged whether the maximum face matching degree is greater than the threshold to perform face recognition, but more factors are considered in the face recognition process, so that the result of face recognition can be made More accurate, reducing the risk of misidentification.

It should be noted that the face recognition method can be applied to various scenarios such as payment scenarios, security monitoring, and attendance check-in, which are not limited here.

The following describes the face recognition method provided in this application by taking a payment scenario as an example.

In the process of using the electronic device for payment, the electronic device first collects the first face image of the target recognition object, and sends the first face image to the payment server (ie, the face recognition device).

After the payment server receives the first face image, based on the above-mentioned face recognition method 100, first according to the first face image and the face images of each predetermined object in the first face database, determine the target recognition object and each predetermined object face matching degree; then obtain the first M face matching degree; then, calculate the difference between the matching degrees of every two adjacent faces, and determine the first object matching the face of the target recognition object in a plurality of predetermined objects Finally, according to the result that the target recognition object and the face of the first object are successfully matched, the payment process is performed, and after the payment process is performed, the successful payment result is returned to the electronic device.

After receiving the successful payment result sent by the payment server, the electronic device displays the payment successful information.

It should also be noted that among each face matching degree, there is only the largest face matching degree greater than the first predetermined threshold, and the other face matching degrees are all below the first predetermined threshold, indicating the recognition result of this time. It is relatively reliable, and it can be directly determined that the predetermined object corresponding to the maximum face matching degree is successfully matched with the face of the target recognition object.

In one or more embodiments of the present application, S110 may include:

In the case that the M-1 difference values satisfy the second predetermined condition, the predetermined object corresponding to the maximum face matching degree is determined as the first object, wherein the second predetermined condition includes: the first object in the M-1 difference values A difference is greater than a predetermined difference threshold, and the first difference is the difference between the largest face matching degree and the second largest human face matching degree.

As some examples, the predetermined difference threshold may be determined according to multiple positive samples for which face recognition is successful, where the multiple positive samples are multiple face image samples for which face recognition is successful.

In the embodiment of the present application, the predetermined difference threshold is determined according to a plurality of positive samples of successful face recognition, which may specifically include:

Perform the following steps for each face image sample: obtaining the similarity between the face image sample and the predetermined object in the first face database; arranging each similarity in descending order; obtaining the first similarity and The second-ranked similarity; calculate the similarity difference between the first-ranked similarity and the second-ranked similarity, so that the similarity difference corresponding to the face image sample is obtained;

After obtaining the similarity difference values corresponding to each face image sample, the similarity difference value with the largest percentage may be counted among the similarity difference values, and the similarity difference value with the largest percentage may be determined as a predetermined difference threshold. For example, if the similarity difference corresponding to 80% of the face image samples is 5 points, 5 is determined as the predetermined difference threshold.

Of course, the method of determining the predetermined difference threshold is not limited to the above method, and after obtaining the similarity difference corresponding to each face image sample, the average value of the similarity difference corresponding to each face image sample can be calculated, A predetermined difference threshold is obtained.

Due to the analysis of the face matching degree, it is concluded that in the samples with successful face matching, the largest face matching degree (ie the first face matching degree) is higher than the second largest face matching degree (ie the second person face matching) is much larger. Therefore, if the difference between the largest face matching degree and the next largest face matching degree is greater than the predetermined difference threshold, it can be considered that the face recognition result is more credible, and the largest face matching degree is determined. The predetermined object corresponding to the degree matches the first object. As a result, the face recognition scheme is optimized, so that the result of face recognition is more accurate, and the risk of misrecognition is reduced.

In one or more embodiments of the present application, the second predetermined condition may further include: the second difference value is not greater than the predetermined difference value threshold, and the second difference value is divided by the first difference value among the M-1 difference values outside difference.

In this embodiment of the present application, the second predetermined condition includes: the first difference is greater than the predetermined difference threshold, and the remaining second differences are not greater than the predetermined difference threshold, that is to say, the largest face matching degree is the second largest person The difference between the face matching degrees is relatively large, and the difference between the remaining face matching degrees is relatively small, that is, the overall face matching degree sequence shows a trend of "steep first, smooth subsequent". In this case, the identification result of this time is considered to be credible. In this way, the accuracy of the face recognition result can be further guaranteed.

The embodiments of the present application are described below through specific examples.

After determining the face matching degree between the target recognition object and each predetermined object, if the face matching degrees between the target recognition object and multiple predetermined objects are all greater than the first predetermined threshold, the first 5 faces are obtained from the face matching degree sequence. Matching degree, 5 face matching degrees are arranged in descending order.

Among the five face matching degrees, if the difference between the first human face matching degree and the second human face matching degree is greater than the predetermined difference threshold, and the difference between the second human face matching degree and the third human face matching degree, The difference between the third face matching degree and the fourth face matching degree, and the difference between the fourth face matching degree and the fifth face matching degree are respectively smaller than the predetermined difference thresholds, that is, the five face matching degrees show the "header" Steep, follow-up smooth" change trend, the recognition result is considered credible, and the predetermined object corresponding to the maximum face matching degree is determined to match the target recognition object, otherwise the recognition is considered to be pending.

After determining the face matching degree between the target recognition object and each predetermined object, if the difference between the maximum face matching degree and the first predetermined threshold is less than the predetermined value, the first two face matching degrees are obtained from the face matching degree sequence. , when the difference between the two face matching degrees is greater than the predetermined difference threshold, the recognition result of this time is considered to be credible, and it is determined that the predetermined object corresponding to the maximum face matching degree matches the target recognition object, otherwise it is considered that this time Identification pending.

For the situation that the identification is pending this time, different auxiliary business methods can be carried out in cooperation with the business, such as requiring the user to input additional identity verification information, which includes at least one of the following: registered mobile phone number, ID number, etc. Of course, if the identification is pending this time, it can also be rejected directly, prompting the user that the identification cannot be made this time.

In the related art, a face image of each predetermined object is pre-stored in the first face database, and the face matching degree between the target recognition object and each predetermined object is calculated by comparing a single image. Specifically, the similarity between the face image of the target recognition object and the face image of the predetermined object is determined as the face matching degree.

However, due to the influence of the shooting environment and shooting angle, sometimes a similarity between the face image of the target recognition object and the face image of the predetermined object cannot accurately determine the face matching degree between the two. In this way, the face recognition result will be inaccurate.

In order to solve the technical problem that the face recognition result is not accurate enough because the face matching degree between the target recognition object and the predetermined object cannot be accurately determined, in one or more embodiments of the present application, the first face database includes N face images of the same predetermined object, where N is an integer greater than 1.

In this embodiment of the present application, as shown in FIG. 2 , S104 may include:

S1042, for N face images of the same predetermined object in the first face database, calculate the similarity between the first face image and each face image in the N face images respectively, and obtain N similarities;

S1044, according to the N similarities, determine the face matching degree between the target recognition object and the predetermined object.

In S1042, the Euclidean distance between the first face image and each of the N face images can be calculated to obtain the similarity; or, the difference between the first face image and each of the N face images can be calculated. Cosine distance to get similarity.

In S1044, S1044 may specifically include one of the following:

Determine the maximum similarity among the N similarities as the face matching degree between the target recognition object and the predetermined object;

In the case where the first similarity among the N similarities is greater than the first predetermined threshold, the first similarity is determined as the face matching degree between the target recognition object and the predetermined object, and the first similarity is the first face image and the predetermined object. similarity of the second face image in the plurality of face images, where the second face image is a face image collected during face recognition setting;

The average value of at least part of the similarity degrees in the N similarities is determined as the face matching degree between the target recognition object and the predetermined object, and specifically, the average value of the N similarities can be determined as the target recognition object and the predetermined object. Face matching degree, or, multiple similarities greater than a certain threshold can be obtained from the N similarities, and the average value of the multiple similarities can be calculated to obtain the face matching degree;

According to the weight value of each similarity in the N similarities, weighted calculation is performed on the N similarities to obtain the face matching degree between the target recognition object and the predetermined object.

The weighted calculation of the N similarities is exemplarily described below.

As some examples, the first face database includes N face images of predetermined objects, and the N face images include a second face image collected during the face recognition setting, and during the face recognition process Added N-1 face images.

Assuming that the weight value of the similarity between the first face image and the second face image is a1, a1 ∈ [0, 1], then, the weight value of the similarity between the first face image and other face images

i is an integer, and i∈[2,N].

The face matching degree between the target recognition object and the predetermined object is: P1×a1+P2×a2+...PN×aN, Pi represents the similarity between the first face image and the ith face image in the N face images.

Since the second face image collected during face recognition setting is more explanatory to a certain extent, when calculating the face matching degree, the similarity between the first face image and the second face image can be has a larger weight value.

It should be noted that the quality of the second face image collected during the face recognition setting is relatively high, specifically, the face angle, face size, face occlusion, closed eyes, Sharpness, pixel size and exposure parameters are within a certain range of parameters.

In the embodiment of the present application, the first face database stores multiple face images of the same predetermined object, and the similarity between the first face image and each face image of the predetermined object is calculated to obtain a plurality of similarities, according to The multiple similarity degrees can accurately determine the face matching degree between the target recognition object and the predetermined object, so that the face recognition can be performed more accurately.

In one or more embodiments of the present application, after S110, the face recognition method may further include:

In the case that the maximum face matching degree is greater than the second predetermined threshold, the first face image is added to the first face database as the face image of the first object, wherein the second predetermined threshold is greater than the first predetermined threshold The value of the threshold.

As some examples, the conditions for adding the first face image to the first face database must not only satisfy that the maximum face matching degree is greater than the second predetermined threshold, but also satisfy the following conditions: the face angle of the first face image, the Face size, face occlusion, closed eyes, clarity, pixel size, and exposure parameters are within a certain range of parameters, and after determining the object matching the target recognition object, the business using face recognition (such as payment business) )Completed successfully.

In this way, the first face database can include multiple face images of the same predetermined object.

In one or more embodiments of the present application, after adding the first face image as the face image of the first object to the first face database, the face recognition method 100 may further include:

In the case that the number of face images of the first object in the first face database is greater than the predetermined number threshold, a third face image is obtained from the face images of the first object in the first face database, and the third face The image is a face image other than the face image collected during the face recognition setting;

According to the face matching degrees corresponding to the respective third human face images, the face image with the smallest face matching degree in at least one third human face image is deleted.

In the embodiment of the present application, the face matching degree corresponding to the third face image is: the object in the third face image obtained by recognizing the face in the third face image and the person of the first object face matching.

It should be noted that, in the case that the number of face images of the first object in the first face database is not greater than the predetermined number threshold, the face images of the first object in the first face database do not need to be deleted.

The embodiments of the present application are exemplarily described below.

It is assumed that there are two face images of the first object in the first face database, which are face image 1 and face image 2 respectively. The face image 1 is a face image collected during face recognition setting, and the face image 2 is a face image added to the first face database during the face recognition process.

When performing face recognition, according to the face image 3 of the target recognition object and the two face images of the first object, determine the face matching degree Y1 between the target recognition object and the first object, and determine the target recognition object and the first object. face matching. In this case, the face image 3 may be added to the first face database as the face image of the first object. In this way, the first face database has three face images of the first object.

When performing face recognition after that, according to the face image 4 of the target recognition object and the three face images of the first object, determine the face matching degree Y2 between the target recognition object and the first object, and determine the target recognition object and the first object. Face matching of an object. In this case, the face image with the face image 4 as the first object may be added to the first face database.

Since the number of face images of the first object in the first face database is greater than the predetermined number threshold of 3, it is necessary to delete a face image of the first object, so that the face images of the first object in the first face database The number is equal to the predetermined number threshold of three. Specifically, first obtain the face matching degrees corresponding to each of the third face images of the first object in the first face database (that is, face image 2 to face image 4), wherein the person corresponding to face image 3 The face matching degree is the above-mentioned Y1, and the face matching degree corresponding to the face image 4 is the above-mentioned Y1. Among face images 2 to 4, the face image with the smallest face matching degree is deleted.

In this way, face images with poor quality in the first face database are deleted. Moreover, by continuously adding face images to the first face database and continuously deleting face images of poor quality in the face database, the face images in the first face database are continuously updated, and the better-quality face images are saved. face image. When using the updated first face database for face recognition, the accuracy of the face recognition result can be guaranteed.

It should be noted that the face image collected during face recognition setting (such as face image 1 in the above example) does not participate in the update of the first face database, but the face image ( For example, face image 2 to face image 4) in the above example participate in the update of the first face database.

In one or more embodiments of the present application, as shown in FIG. 3 , after S110, the face recognition method 100 may further include:

S112, in the face image of the first object in the first face database, determine whether there is a fourth face image whose similarity with the first face image is less than a third predetermined threshold, where the third predetermined threshold is less than the first the value of the predetermined threshold;

S114, in the presence of the fourth human face image, add 1 to the first parameter value corresponding to the fourth human face image, where the first parameter value represents the fourth human face image and the first human face image of the at least one target recognition object The number of mismatches between;

S116, in the case that the first parameter value is greater than the predetermined first number of times threshold, delete the fourth human face image from the first human face database.

In this embodiment of the present application, if the target recognition object matches the first object, the similarity between the first face image of the target recognition object and the fourth face image of the first object in the first face database is small. In this case, the value of the first parameter corresponding to the fourth face image is increased by 1, that is, the number of mismatches of the fourth face image is increased by 1. In the case where the value of the first parameter is greater than the predetermined threshold of the first number of times, it may be because the face matching of the fourth face image is too high due to individual angles and other reasons in the face recognition process, which leads to the fourth The face image was mistakenly added to the first face database. At this time, the fourth face image can be deleted from the first face database, so that the face image added to the first face database by mistake can be deleted, so as to continuously optimize the first face database.

In one or more embodiments of the present application, the face recognition method 100 may further include: in the case that the M-1 difference values do not satisfy the second predetermined condition, obtaining the identity verification information (such as the input of the target recognition object) ID number or verification code), and identify the target identification object according to the identity verification information.

In one or more embodiments of the present application, the above-mentioned first face image and the face images of each predetermined object in the first face database may be two-dimensional Red Green Blue (RGB) face image. In the case where the object matching the target recognition object cannot be recognized according to the two-dimensional face image, since the three-dimensional face image contains the height information of each part of the face, the second person including the three-dimensional face image can be face database for recognition.

Specifically, the second face database can be pre-built. Wherein, three-dimensional face images may be collected when the user performs face recognition settings, so as to construct a second face database. However, the camera hardware of the electronic device is different. If the hardware of the electronic device does not support the acquisition of three-dimensional face images, the three-dimensional face images of the predetermined objects in the second face database are empty. If the electronic device collects the three-dimensional face image of the predetermined object, and the collected three-dimensional face image meets the quality requirements, the collected three-dimensional face image is saved to the second face database.

In addition, in face-swiping scenarios such as face-swiping payment, access control, etc., when electronic devices perform face recognition, they need to perform liveness detection, often with specific hardware cameras that can support the collection of three-dimensional face images. After the recognized object passes the live detection, the recognition is successful, and the business is completed (such as successful payment or access control unlocking), the collected three-dimensional face image is added to the second face database of the recognized object.

It should be noted that each three-dimensional face image in the second face database corresponds to the second parameter value 3D_OKR one-to-one. The second parameter value 3D_OKR corresponding to the three-dimensional face image represents the number of times of successful face recognition using the three-dimensional face image. The initial value of the second parameter value 3D_OKR corresponding to the three-dimensional face image is 0.

If the second parameter value 3D_OKR corresponding to a 3D face image is greater than the predetermined second time threshold POSITIVE_MAX_3DOKR, it means that the 3D face image is used for more successful face recognition, and the 3D face image is credible of. Therefore, if the recognition is successful according to the 3D face image, it is not necessary to additionally obtain the identity verification information input by the target recognition object for recognition, but the recognition result of the 3D face image can be used.

In this embodiment of the present application, when the second time threshold POSITIVE_MAX_3DOKR takes a value of 0, it means that the step of obtaining the identity verification information input by the target recognition object for recognition is omitted, and the 3D image-assisted recognition strategy can be directly used.

After the second face database is constructed, if there is a risk of misrecognition when using the first face database for identification, a three-dimensional image-assisted identification strategy is started. Specifically, as shown in FIG. 4, after S110, the face recognition method 100 may further include:

S118, in the case that the M-1 differences do not meet the second predetermined condition, obtain a first three-dimensional face image of the target recognition object;

S120, matching the first three-dimensional face image with the three-dimensional face image of the predetermined object in the second face database;

S122, the first three-dimensional face image is successfully matched with the second three-dimensional face image of the second object in the second face database, and the second parameter value corresponding to the second three-dimensional face image is greater than the predetermined second time threshold. In this case, it is determined that the target recognition object matches the face of the second object.

It should be noted that the second object and the first object may be the same object, or may be different objects.

In the embodiment of the present application, the sensorless verification can be performed assisted by the three-dimensional face image in the second face database, so as to reduce the user's input and perception, and improve the user experience.

In one or more embodiments of the present application, after S110, the face recognition method 100 may further include:

When the second parameter value 3D_OKR is not greater than the second threshold POSITIVE_MAX_3DOKR, obtain the identity verification information of the target recognition object, and the identity verification information is information other than face information;

Use the identity verification information to identify the identity, and obtain the identification result;

In the case that the recognition result using the identity verification information is consistent with the face recognition result using the first three-dimensional face image, add 1 to the second parameter value 3D_OKR;

When the face recognition result using the first three-dimensional face image is inconsistent with the recognition result using the identity verification information, the second parameter value 3D_OKR is decreased by 1.

In the embodiment of the present application, the recognition result using the identity verification information is consistent with the face recognition result using the first three-dimensional face image, which may specifically include:

The identity verification information is successfully matched with the predetermined identity information of the second object, and the first three-dimensional face image is successfully matched with the second three-dimensional face image of the second object in the second face database;

Alternatively, the identity verification information does not match the predetermined identity information of the second object, and the first three-dimensional face image does not match the second three-dimensional face image of the second object in the second face database.

The face recognition result using the first three-dimensional face image is inconsistent with the recognition result using the identity verification information, which may specifically include:

The identity verification information does not match the predetermined identity information of the second object, and the first three-dimensional face image is successfully matched with the second three-dimensional face image of the second object in the second face database;

Or, the identity verification information is successfully matched with the predetermined identity information of the second object, and the first three-dimensional face image does not match the second three-dimensional face image of the second object in the second face database.

In one or more embodiments of the present application, after subtracting 1 from the second parameter value 3D_OKR, the face recognition method 100 may further include:

In the case where the second parameter value 3D_OKR after subtracting 1 is smaller than the predetermined third threshold of times NEGATIVE_MAX_3DOKR, it means that the number of times of face recognition failure by using the second three-dimensional face image is relatively high, then delete the third face database in the second face database. 2D and 3D face images. In this way, the second face database is continuously updated, thereby continuously improving the quality of the three-dimensional face images in the second face database, and reducing the risk of using the three-dimensional face images in the second face database for face recognition.

In this embodiment of the present application, the third time threshold NEGATIVE_MAX_3DOKR is smaller than the second time threshold POSITIVE_MAX_3DOKR.

Corresponding to the face recognition method provided by the present application, the present application also provides a face recognition device. FIG. 5 is a schematic structural diagram of an embodiment of a face recognition device provided by the present application. As shown in Figure 5, the face recognition device 200 includes:

The first obtaining module 202 is used for obtaining the first face image of the target recognition object;

The first determination module 204 is used to determine the face matching degree between the target recognition object and each predetermined object according to the first face image and the face image of each predetermined object in the first face database;

The second obtaining module 206 is configured to obtain the first M face matching degrees from the face matching degree sequence when the face matching degree satisfies the first predetermined condition, and the face matching degree sequence is the target recognition object and each predetermined object M is an integer greater than 1, and the first predetermined condition includes at least one of the following: the face matching degree of the target recognition object and a plurality of predetermined objects is greater than the first a predetermined threshold, the difference between the maximum face matching degree and the first predetermined threshold is less than a predetermined value;

The first calculation module 208 is used to calculate the difference between the matching degrees of each adjacent two faces in the M face matching degrees, and obtain M-1 difference values;

The second determining module 210 is configured to determine, according to the M−1 difference values, a first object matching the face of the target recognition object among the plurality of predetermined objects.

In one or more embodiments of the present application, the second determining module 210 may specifically be used to:

Due to the analysis of the face matching degree, it is concluded that among the samples with successful face matching, the largest face matching degree (ie the first face matching degree) is higher than the second largest face matching degree (ie the second face matching degree) match) is much larger. Therefore, if the difference between the largest face matching degree and the next largest face matching degree is greater than the predetermined difference threshold, it can be considered that the face recognition result is more credible, and the largest face matching degree is determined. The predetermined object corresponding to the degree matches the first object. As a result, the face recognition scheme is optimized, so that the result of face recognition is more accurate, and the risk of misrecognition is reduced.

In one or more embodiments of the present application, the second predetermined condition further includes: the second difference value is not greater than a predetermined difference value threshold, and the second difference value is in addition to the first difference value among M-1 difference values difference value.

In one or more embodiments of the present application, the first determining module 204 may include:

The first calculation unit is used to calculate the similarity between the first face image and each face image in the N face images for N face images of the same predetermined object in the first face database, and obtain N similarities degree, N is an integer greater than 1;

The first determining unit is configured to determine, according to the N similarities, the degree of face matching between the target recognition object and the predetermined object.

In one or more embodiments of the present application, the first determining unit may include one of the following:

a first determination subunit, used for determining the maximum similarity among the N similarities as the face matching degree between the target recognition object and the predetermined object;

The second determination subunit is configured to determine the first similarity as the face matching degree between the target recognition object and the predetermined object when the first similarity among the N similarities is greater than the first predetermined threshold, and the first similarity is The degree is the similarity between the first face image and the second face image in the plurality of face images, and the second face image is the face image collected when the face recognition setting is performed;

a third determination subunit, configured to determine the average value of at least part of the similarity among the N similarities as the face matching degree between the target recognition object and the predetermined object;

The fourth determination subunit is configured to perform weighted calculation on the N similarities according to the weight value of each similarity in the N similarities, so as to obtain the face matching degree between the target recognition object and the predetermined object.

In one or more embodiments of the present application, the face recognition apparatus 200 may further include:

The adding module is configured to add the first face image as the face image of the first object to the first face database when the maximum face matching degree is greater than the second predetermined threshold, wherein the second predetermined threshold is a value greater than the first predetermined threshold.

A third acquisition module, configured to acquire a third person from the face images of the first object in the first face database when the number of face images of the first object in the first face database is greater than a predetermined number threshold face image, the third face image is a face image other than the face image collected during face recognition setting;

The first deletion module is used to delete the face image with the smallest face matching degree in at least one third face image according to the face matching degree corresponding to each third face image, wherein the third face image corresponds to the face image. The face matching degree is: the face matching degree between the object in the third face image and the first object obtained by recognizing the face in the third face image.

In the embodiment of the present application, the face images with poor quality in the first face database can be deleted, so that the face images in the first face database are continuously updated, and the face images with better quality are saved. When using the updated first face database for face recognition, the accuracy of the face recognition result can be guaranteed.

The third determining module is configured to determine, in the face image of the first object in the first face database, whether there is a fourth face image whose similarity with the first face image is less than a third predetermined threshold, the third predetermined The threshold is a value smaller than the first predetermined threshold;

The second calculation module is configured to add 1 to the first parameter value corresponding to the fourth human face image in the presence of the fourth human face image, where the first parameter value represents the relationship between the fourth human face image and at least one target recognition object The number of mismatches between the first face images;

The second deletion module is configured to delete the fourth human face image from the first human face database when the value of the first parameter is greater than the predetermined threshold of the first number of times.

a fourth acquisition module, configured to acquire the first three-dimensional face image of the target recognition object when the M-1 differences do not meet the second predetermined condition;

a matching module for matching the first three-dimensional face image with the three-dimensional face image of the predetermined object in the second face database;

The fourth determination module is used to successfully match the first three-dimensional face image with the second three-dimensional face image of the second object in the second face database, and the second parameter value corresponding to the second three-dimensional face image is greater than a predetermined value. In the case of the second number of times threshold, it is determined that the target recognition object matches the face of the second object, wherein the second parameter value represents the number of successful face recognition using the second three-dimensional face image.

a fifth acquisition module, configured to acquire the identity verification information of the target identification object when the second parameter value is not greater than the second number of times threshold;

The identification module is used to identify the identity by using the identity verification information to obtain the identification result;

The third computing module is used to add 1 to the second parameter value when the recognition result using the identity verification information is consistent with the face recognition result using the first three-dimensional face image;

The fourth calculation module is configured to reduce the value of the second parameter by 1 when the face recognition result using the first three-dimensional face image is inconsistent with the recognition result using the identity verification information.

A third deletion module, configured to delete the second three-dimensional face image in the second face database when the second parameter value after subtracting 1 is less than a predetermined third threshold of times, and the The third threshold of times is smaller than the second threshold of times.

In this way, the second face database can be continuously updated, thereby continuously improving the quality of the three-dimensional face images in the second face database, and reducing the risk of using the three-dimensional face images in the second face database for face recognition.

The present application also provides a face recognition device comprising: a processor and a memory storing computer program instructions, the processor implements the steps of any one of the above-mentioned face recognition method embodiments when the processor executes the computer program instructions.

As shown in FIG. 6 , the face recognition device may include a processor 301 and a memory 302 storing computer program instructions.

Specifically, the above-mentioned processor 301 may include a central processing unit (CPU), or a specific integrated circuit (Application Specific Integrated Circuit, ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

Memory 302 may include mass storage for data or instructions. By way of example and not limitation, memory 302 may include a Hard Disk Drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a Universal Serial Bus (USB) drive or two or more A combination of more than one of the above. Memory 302 may include removable or non-removable (or fixed) media, where appropriate. Storage 302 may be internal or external to the integrated gateway disaster recovery device, where appropriate. In certain embodiments, memory 302 is non-volatile solid state memory.

Memory may include read only memory (ROM), random access memory (RAM), magnetic disk storage media devices, optical storage media devices, flash memory devices, electrical, optical or other physical/tangible memory storage devices. Thus, typically, a memory includes one or more tangible (non-transitory) computer-readable storage media (eg, memory devices) encoded with software including computer-executable instructions, and when the software is executed (eg, by a or multiple processors), it is operable to perform the operations described with reference to a method according to an aspect of the present disclosure.

The processor 301 reads and executes the computer program instructions stored in the memory 302 to implement any one of the face recognition methods in the foregoing embodiments.

In some examples, the facial recognition device may also include a communication interface 303 and a bus 310 . Among them, as shown in FIG. 6 , the processor 301 , the memory 302 , and the communication interface 303 are connected through the bus 310 and complete the mutual communication.

The communication interface 303 is mainly used to implement communication between modules, apparatuses, units and/or devices in the embodiments of the present application.

The bus 310 includes hardware, software, or both, coupling the components of the online data flow metering device to each other. By way of example and not limitation, bus 310 may include Accelerated Graphics Port (AGP) or other graphics bus, Enhanced Industry Standard Architecture (EISA) bus, Front Side Bus (FSB), HyperTransport (HT) interconnect, Industry Standard Architecture (ISA) ) bus, Infiniband Interconnect, Low Pin Count (LPC) bus, Memory Bus, Microchannel Architecture (MCA) bus, Peripheral Component Interconnect (PCI) bus, PCI-Express (PCI-X) bus, Serial Advanced Technology Attachment (SATA) bus, Video Electronics Standards Association Local (VLB) bus or other suitable bus or a combination of two or more of the above. Bus 310 may include one or more buses, where appropriate. Although embodiments herein describe and illustrate a particular bus, this application contemplates any suitable bus or interconnect.

In addition, in combination with the face recognition method in the foregoing embodiment, the embodiment of the present application may provide a computer storage medium for implementation. Computer program instructions are stored on the computer storage medium; when the computer program instructions are executed by the processor, any one of the face recognition methods in the foregoing embodiments is implemented. Examples of computer-readable storage media shown include non-transitory computer-readable storage media, such as read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk, etc. .

To be clear, the present application is not limited to the specific configurations and processes described above and illustrated in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above-described embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the sequence of steps after comprehending the spirit of the present application.

The functional blocks shown in the above-described structural block diagrams may be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, it may be, for example, an electronic circuit, an application specific integrated circuit (ASIC), suitable firmware, a plug-in, a function card, or the like. When implemented in software, elements of the present application are programs or code segments used to perform the required tasks. The program or code segments may be stored in a machine-readable medium or transmitted over a transmission medium or communication link by a data signal carried in a carrier wave. A "machine-readable medium" may include any medium that can store or transmit information. Machine-readable media may include non-transitory computer-readable storage media, such as including electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, machine The readable medium may also include a radio frequency (RF) link, among others. The code segments may be downloaded via a computer network such as the Internet, an intranet, or the like.

It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, the present application is not limited to the order of the above steps, that is, the steps may be performed in the order mentioned in the embodiment, or may be different from the order in the embodiment, or several steps may be performed simultaneously.

Aspects of the present application are described above with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the present application. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine such that execution of the instructions via the processor of the computer or other programmable data processing apparatus enables the Implementation of the functions/acts specified in one or more blocks of the flowchart and/or block diagrams. Such processors may be, but are not limited to, general purpose processors, special purpose processors, application specific processors, or field programmable logic circuits. It will also be understood that each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations, can also be implemented by special purpose hardware for performing the specified functions or actions, or by special purpose hardware and/or A combination of computer instructions is implemented.

The above are only specific implementations of the present application. Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working process of the above-described systems, modules and units may refer to the foregoing method embodiments. The corresponding process in , will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should all cover within the scope of protection of this application.

Claims

A face recognition method, comprising:

Obtain the first face image of the target recognition object;

According to the first face image and the face images of each predetermined object in the first face database, determine the face matching degree between the target recognition object and each of the predetermined objects;

In the case that the face matching degree satisfies the first predetermined condition, the first M face matching degrees are obtained from the face matching degree sequence, and the face matching degree sequence is the target recognition object and each of the predetermined objects. M is an integer greater than 1, and the first predetermined condition includes at least one of the following: the target recognition object and a plurality of people of the predetermined object The face matching degree is greater than a first predetermined threshold, and the difference between the maximum face matching degree and the first predetermined threshold is less than a predetermined value;

Calculate the difference value of each adjacent two described human face matching degrees in the M described human face matching degrees, and obtain M-1 difference values;

According to the M-1 difference values, a first object matching the face of the target recognition object is determined from among the plurality of predetermined objects.
The method according to claim 1, wherein, according to the M-1 difference values, determining a first object matching the face of the target recognition object among the plurality of predetermined objects comprises:

In the case that the M-1 difference values satisfy a second predetermined condition, the predetermined object corresponding to the largest face matching degree is determined as the first object, wherein the second predetermined condition includes : the first difference value among the M-1 difference values is greater than a predetermined difference value threshold, and the first difference value is the difference between the maximum matching degree of the human face and the second largest matching degree of the human face value.
The method according to claim 2, wherein the second predetermined condition further comprises: a second difference value is not greater than the predetermined difference value threshold, and the second difference value is one of the M−1 difference values. Differences other than the first difference.
The method according to claim 1, wherein the target recognition object and the person of each predetermined object are determined according to the first face image and the face image of each predetermined object in the first face database Face matching, including:

For N face images of the same predetermined object in the first face database, the similarity between the first face image and each face image in the N face images is calculated, and N face images are obtained. Similarity, N is an integer greater than 1;

According to the N similarities, the face matching degree between the target recognition object and the predetermined object is determined.
The method according to claim 4, wherein the determining, according to the N similarities, the degree of face matching between the target recognition object and the predetermined object, comprising one of the following:

Determining the maximum similarity among the N similarities as the face matching degree between the target recognition object and the predetermined object;

In the case where the first similarity among the N similarities is greater than the first predetermined threshold, the first similarity is determined as the face matching degree between the target recognition object and the predetermined object, and the The first similarity is the similarity between the first face image and the second face image in the plurality of face images, and the second face image is the face collected during the face recognition setting image;

determining the average value of at least part of the similarity among the N similarities as the face matching degree between the target recognition object and the predetermined object;

According to the weight value of each similarity in the N similarities, weighted calculation is performed on the N similarities, so as to obtain the face matching degree between the target recognition object and the predetermined object.
The method according to claim 1, wherein after the object corresponding to the target recognition object is determined from the plurality of predetermined objects according to the M-1 difference values, the method further comprises:

In the case that the maximum matching degree of the face is greater than the second predetermined threshold, the first face image is added to the first face database as the face image of the first object, wherein the The second predetermined threshold is a value greater than the first predetermined threshold.
The method according to claim 6, wherein after adding the first face image as the face image of the first object to the first face database, the method further comprises:

In the case where the number of face images of the first object in the first face database is greater than a predetermined number threshold, obtain a third a face image, the third face image is a face image other than the face image collected during the face recognition setting;

Delete at least one face image with the smallest face matching degree among the third face images according to the face matching degrees corresponding to each of the third face images, wherein the person corresponding to the third face image is deleted. The face matching degree is: the face matching degree between the object in the third human face image and the first object obtained by recognizing the human face in the third human face image.
The method according to claim 4, wherein after the first object corresponding to the target identification object is determined from the plurality of predetermined objects according to the M-1 difference values, the method further comprises:

In the face image of the first object in the first face database, it is determined whether there is a fourth face image whose similarity with the first face image is less than a third predetermined threshold, the third face image The predetermined threshold is a value smaller than the first predetermined threshold;

In the case where the fourth human face image exists, add 1 to the first parameter value corresponding to the fourth human face image, where the first parameter value indicates that the fourth human face image and at least one of the targets The number of mismatches between the first face images of the identified object;

In the case that the value of the first parameter is greater than a predetermined threshold of the first number of times, the fourth face image is deleted from the first face database.
The method according to claim 2, wherein after calculating the difference between each adjacent two of the M face matching degrees, and obtaining M-1 difference values, the Methods also include:

In the case that the M-1 difference values do not meet the second predetermined condition, acquiring a first three-dimensional face image of the target recognition object;

matching the first three-dimensional face image with the three-dimensional face image of the predetermined object in the second face database;

The first three-dimensional face image is successfully matched with the second three-dimensional face image of the second object in the second face database, and the second parameter value corresponding to the second three-dimensional face image is greater than a predetermined first In the case of a quadratic number threshold, it is determined that the target recognition object matches the face of the second object, wherein the second parameter value represents the number of times of successful face recognition using the second three-dimensional face image.
The method according to claim 9, characterized in that, after calculating the difference between each adjacent two of the M face matching degrees, after obtaining M-1 difference values, The method also includes:

Under the condition that the second parameter value is not greater than the second threshold of times, obtain the identity verification information of the target identification object;

Use the identity verification information to perform identity recognition to obtain the recognition result;

When the recognition result using the identity verification information is consistent with the face recognition result using the first three-dimensional face image, add 1 to the second parameter value;

In the case that the face recognition result using the first three-dimensional face image is inconsistent with the recognition result using the identity verification information, the second parameter value is decreased by 1.
The method according to claim 10, wherein after the decreasing the second parameter value by 1, the method further comprises:

In the case that the second parameter value after subtracting 1 is smaller than a predetermined third time threshold, delete the second three-dimensional face image in the second face database, and the third time threshold is smaller than the third time threshold Second time threshold.
A face recognition device, comprising:

a first acquisition module, used for acquiring the first face image of the target recognition object;

a first determination module, configured to determine the face matching degree between the target recognition object and each of the predetermined objects according to the first face image and the face images of each predetermined object in the first face database;

The second obtaining module is configured to obtain the first M face matching degrees from the face matching degree sequence when the face matching degree satisfies the first predetermined condition, and the face matching degree sequence is the target recognition A sequence obtained by arranging the face matching degrees of the objects and each of the predetermined objects in descending order; M is an integer greater than 1, and the first predetermined condition includes at least one of the following: the target recognition object and multiple The face matching degree of each of the predetermined objects is greater than a first predetermined threshold, and the difference between the largest face matching degree and the first predetermined threshold is less than a predetermined value;

The first calculation module is used to calculate the difference between each adjacent two described face matching degrees in the M described face matching degrees, and obtain M-1 difference values;

The second determining module is configured to determine, according to the M-1 difference values, a first object matching the face of the target recognition object from among the plurality of predetermined objects.
A face recognition device, comprising: a processor and a memory storing computer program instructions;

The processor implements the steps of the face recognition method according to any one of claims 1-11 when the processor executes the computer program instructions.
A computer storage medium, where computer program instructions are stored thereon, and when the computer program instructions are executed by a processor, the steps of the face recognition method according to any one of claims 1-11 are implemented.