Method, apparatus and electronic device for updating biometric database

Abstract

The present disclosure provides a method, an apparatus, an electronic device, and a computer-readable storage medium for updating a biometric database. The biometric database includes at least one reference biometric. The method for updating the biometric database includes: obtaining a similarity between the input biometric and a target reference biometric, where the input biometric is a biometric extracted from multimedia data, and the target reference biometric is the reference biometric with the highest identity matching degree with the input biometric among the at least one reference biometric; when the similarity is greater than a first threshold, performing feature fusion on the input biometric and the target reference biometric to obtain a fused feature; and updating the target reference biometric to the fused feature.

Description

[0001] This application claims priority to a Chinese patent application filed with the China Patent Office on October 20, 2020, with application number 202011124636.9 and invention name “Method, device and electronic device for updating biometric library”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present disclosure relates to the field of artificial intelligence technology, and in particular to a method, device, electronic device and computer-readable storage medium for updating a biometric database. Background Art

[0003] In today's information age, how to accurately identify individuals and protect information security has become a key social issue that must be resolved. Biometric-based identity recognition technology (referred to as biometric recognition) refers to the technology of using a computer to identify an individual using the physiological or behavioral characteristics inherent in an organism. Among the human body feature-based identity recognition technologies, face recognition is widely used due to its convenience and speed.

[0004] Taking face recognition as an example, the features of the face to be recognized are generally compared with the features of the reference face, and the identity recognition result is obtained based on the similarity. However, facial features may change significantly with age, hairstyle, weather, season, etc. Therefore, how to improve the accuracy of face recognition is a technical problem that needs to be solved urgently. Summary of the invention

[0005] The embodiments of the present disclosure provide a method, device, electronic device, and computer-readable storage medium for updating a biometric database to improve the accuracy of individual identity recognition using biometrics.

[0006] According to one aspect of an embodiment of the present disclosure, a method for updating a biometric database is provided, wherein the biometric database includes at least one reference biometric, and the method includes:

[0007] Obtaining a similarity between the input biometric feature and a target reference biometric feature, wherein the input biometric feature is a biometric feature extracted from multimedia data, and the target reference biometric feature is a reference biometric feature with the greatest identity matching degree with the input biometric feature among the at least one reference biometric feature;

[0008] When the similarity is greater than a first threshold, performing feature fusion on the input biometric feature and the target reference biometric feature to obtain a fused feature;

[0009] The target reference biometric feature is updated to the fused feature.

[0010] In some embodiments, the input biometric feature is obtained by performing biometric feature extraction on a single picture containing biometric features; or

[0011] the input biometric feature is obtained by performing biometric feature extraction on multiple pictures containing biometric features and then performing biometric feature fusion.

[0012] In some embodiments, updating the target reference biometric feature to the fused feature includes:

[0013] Searching for the target reference biometric feature in the biometric feature library according to the identity identifier of the target reference biometric feature;

[0014] Replacing the target reference biometric feature with the fused feature.

[0015] In some embodiments, obtaining the similarity of the comparison between the input biometric feature and the target reference biometric feature includes: after completing identity recognition based on the input biometric feature, obtaining the similarity of the comparison between the input biometric feature and the target reference biometric feature;

[0016] The first threshold is greater than the second threshold, and the second threshold is: when performing identity recognition based on the input biometric feature, the minimum similarity between the input biometric feature indicating successful identity recognition and the reference biometric feature.

[0017] In some embodiments, the method for updating the biometric feature library further includes: before obtaining the similarity of the comparison between the input biometric feature and the target reference biometric feature, determining the target reference biometric feature according to the similarity of the comparison between the input biometric feature and the reference biometric feature, and the result information of whether the association attributes between the input biometric feature and the reference biometric feature are consistent;

[0018] wherein, the confidence level of the similarity of the comparison between the input biometric feature and the reference biometric feature is not less than the confidence level threshold, and the association attributes include at least one of gender, age group, and physical constitution type.

[0019] In some embodiments, performing feature fusion on the input biometric feature and the target reference biometric feature to obtain a fused feature includes:

[0020] Performing feature fusion on the input biometric feature and the target reference biometric feature according to the function feat_out = feat_cap*(1–momentum)+feat_old*momentum;

[0021] Among them, feat_out is the fused feature, feat_cap is the input biometric feature, feat_old is the target reference biometric feature, momentum is the momentum coefficient, and 0 ≤ momentum < 1.

[0022] In some embodiments, 0.9 ≤ momentum < 1.

[0023] In some embodiments, the biometric database includes a face feature database, a palmprint feature database, a skin feature database, an auricle feature database, a gait feature database, or a voice feature database.

[0024] According to another aspect of the embodiments of the present disclosure, there is provided a device for updating a biometric database. The biometric database includes at least one reference biometric feature. The device includes:

[0025] An acquisition unit configured to acquire the similarity between the input biometric feature and the target reference biometric feature. The input biometric feature is a biometric feature extracted from multimedia data, and the target reference biometric feature is the reference biometric feature with the highest identity matching degree with the input biometric feature among the at least one reference biometric feature;

[0026] A fusion unit configured to perform feature fusion on the input biometric feature and the target reference biometric feature when the similarity is greater than a first threshold to obtain a fused feature;

[0027] An update unit configured to update the target reference biometric feature to the fused feature.

[0028] According to yet another aspect of the embodiments of the present disclosure, there is provided an electronic device including a memory and a processor coupled to the memory. The processor is configured to execute the method for updating a biometric database according to any one of the foregoing technical solutions based on instructions stored in the memory.

[0029] According to still another aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for updating a biometric database according to any one of the foregoing technical solutions is implemented.

[0030] The above embodiments of the present disclosure can automatically update the biometric database according to program settings. Compared with the fixed reference biometric features in the related art, the reference biometric features in the embodiments of the present disclosure can be dynamically adjusted following the changes of individual biometric features. The updated biometric database can be applied to biometric recognition to improve the recognition accuracy and thus enhance the information security.

[0031] Of course, the products or methods implementing any embodiment of the present disclosure do not necessarily need to achieve all of the above advantages at the same time. Description of the Drawings

[0032] To more clearly illustrate the technical solutions in the embodiments of the present disclosure or related technologies, the following briefly introduces the drawings required for use in the description of the embodiments of the present disclosure or related technologies. Obviously, the drawings in the following description are only some embodiments of the present disclosure. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0033] Figure 1 It is a flowchart of a face recognition method in related technologies;

[0034] Figure 2 It is a flowchart of a method for updating a biometric database in some embodiments of the present disclosure;

[0035] Figure 3 It is a flowchart of face recognition and updating a face feature database in some embodiments of the present disclosure;

[0036] Figure 4 It is a block diagram of a device for updating a biometric database in some embodiments of the present disclosure;

[0037] Figure 5 It is a block diagram of an electronic device in some embodiments of the present disclosure. Detailed Description of the Embodiments

[0038] Next, the technical solutions in the embodiments of the present disclosure will be clearly and completely described in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only some embodiments of the present disclosure, rather than all embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those of ordinary skill in the art based on this application belong to the scope of protection of the present disclosure.

[0039] Face recognition is a biometric recognition technology that performs identity recognition based on the facial features of a person. Its basic principle is to extract facial features from a picture or video frame containing a face area, compare the facial features with the reference facial features pre-established in the facial feature database, and obtain the recognition result of the facial identity information according to the degree of similarity.

[0040] As Figure 1 shown, it is a face recognition method in related technologies, and this method includes the following steps S01 - S05.

[0041] In step S01, obtain a picture containing a face area collected by an image acquisition device.

[0042] In step S02, use a facial feature extraction algorithm to extract facial features from the picture containing the face area as the input facial features.

[0043] In step S03, traverse the reference biometric features in the biometric feature library, compare the input biometric features with each reference biometric feature, and obtain a similarity list.

[0044] In step S04, filter the similarity list, and retain the similarity comparison data with a confidence level not less than the confidence level threshold as the filtering result.

[0045] In step S05, output the recognition result according to the reference biometric feature corresponding to the maximum similarity in the filtering result. The recognition result may include, for example, the maximum similarity value, the face image restored based on the reference biometric feature, the identity identification number of the individual corresponding to the reference biometric feature, and so on.

[0046] The inventors of the present disclosure found in the process of implementing the embodiments of the present disclosure that, since the reference biometric features in the biometric feature library are fixed, when the biometric features change greatly with age, hairstyle, weather, season, etc., the accuracy of the face recognition result will be greatly reduced, thus bringing information security risks.

[0047] To improve the accuracy of individual identity recognition using biometric features, embodiments of the present disclosure provide a method, apparatus, electronic device, and computer-readable storage medium for updating a biometric feature library.

[0048] The technical solutions provided by the embodiments of the present disclosure can be applied to various scenarios that require biometric recognition. For example, security inspection monitoring, public place monitoring, access control monitoring, criminal investigation detection, livestock monitoring, etc. The applicable fields include, but are not limited to, security, finance, border, customs, insurance, and civilian entertainment. According to the different specific application scenarios of biometric recognition, the biometric feature library may be, for example, a face feature library, a palm feature library, a skin feature library, an auricle feature library, a gait feature library, or a voice feature library, etc. The biometric feature library may include only one reference biometric feature, or may include multiple reference biometric features, and the number of reference biometric features in the biometric feature library can be increased or decreased according to user needs.

[0049] As Figure 2 shown, some embodiments of the present disclosure provide a method for updating a biometric feature library, where the biometric feature library includes at least one reference biometric feature, and the method for updating the biometric feature library includes the following steps S1 to S3.

[0050] In step S1, obtain the similarity between the input biometric feature and the target reference biometric feature, where the input biometric feature is the biometric feature extracted from the multimedia data, and the target reference biometric feature is the reference biometric feature with the largest identity matching degree with the input biometric feature among the foregoing at least one reference biometric feature.

[0051] In some embodiments of the present disclosure, the input biometric feature is obtained by performing biometric feature extraction on a single picture containing biometric features. For example, the input biometric feature is extracted from a single photo containing biometric features taken by a camera, or is extracted from one frame of a video stream containing biometric features captured by a video camera.

[0052] In some other embodiments of the present disclosure, the input biometric feature is obtained by performing biometric feature fusion after performing biometric feature extraction on multiple pictures containing biometric features. For example, biometric feature extraction is performed separately on multiple frames of a video stream containing biometric features captured by a video camera, and then these biometric features are fused to obtain the input biometric feature. In this way, when one or some of the picture frames are blurred, have low resolution, or have an unreasonable shooting angle, through biometric fusion calculation, its impact on the accuracy of the recognition result can be made smaller or even negligible, thereby improving the accuracy of biometric recognition.

[0053] In the embodiments of the present disclosure, the reference biometric feature of the biometric database is an offline reference template obtained based on a deep learning algorithm. Deep learning is a type of machine learning, and machine learning is an essential path to achieve artificial intelligence. The concept of deep learning originates from the research of artificial neural networks. A multi-layer perceptron with multiple hidden layers is a type of deep learning structure. Deep learning forms more abstract high-level representations of attribute categories or features by combining low-level features to discover the distributed feature representations of data. The motivation for studying deep learning is to establish a neural network that simulates the human brain for analysis and learning, which mimics the mechanism of the human brain to interpret data, such as images, sounds, and texts. Typical deep learning models include convolutional neural network models, deep belief network models, and stacked autoencoder network models, etc.

[0054] In the embodiments of the present disclosure, the target reference biometric feature refers to: comparing each reference biometric feature in the biometric database with the input biometric feature, and the reference biometric feature with the largest identity matching degree with the input biometric feature. Among them, the factors considered for the identity matching degree include the similarity of biometric feature comparison. In addition, it can also include the consistency of other attributes associated with the user identity, such as the consistency of gender, age group, physical form category, etc. In the biometric database, the attributes associated with the user identity of the reference biometric feature (hereinafter referred to as associated attributes) can be pre-calculated by using a certain algorithm and stored in the memory. When performing biometric recognition, the associated attributes of the input biometric feature can be calculated by using a certain algorithm.

[0055] In some embodiments of the present disclosure, the method for updating the biometric database further includes: before obtaining the similarity between the input biometric and the target reference biometric, determining the target reference biometric according to the similarity between the input biometric and the reference biometric, and the result information on whether the association attributes between the input biometric and the reference biometric are consistent; wherein, the confidence level of the similarity between the input biometric and the reference biometric is not less than the confidence level threshold, and the association attributes include at least one of gender, age group, and physical form category.

[0056] In some embodiments of the present disclosure, the above step S1 is executed after the identity recognition based on the input biometric is completed. In some other embodiments of the present disclosure, the above step S1 may also be executed during the process of identity recognition based on the input biometric, or before the identity recognition based on the input biometric.

[0057] After the identity recognition for the input biometric is completed, generally, the recognition result will be output. The recognition result may include the identity verification determination result of the input biometric. In addition, it may also include the similarity between the input biometric and the target reference biometric, the identity identifier of the individual corresponding to the target reference biometric, the individual image restored based on the target reference biometric, and other association attributes, such as gender, age group, physical form category, etc.

[0058] In some embodiments of the present disclosure, after each identity recognition based on the input biometric is completed, the similarity between the input biometric and the target reference biometric is obtained. That is, every time a biometric recognition program is executed, after the recognition is completed, a biometric database update program is started.

[0059] In some other embodiments of the present disclosure, it is also possible to obtain, every time a set time period elapses, the recognition result output during the identity recognition based on the input biometric within the set time period, and the recognition result includes the similarity between the input biometric and the target reference biometric. That is, the biometric database update program is started at a predetermined frequency. The set time period can be determined in combination with the system processing performance and the update requirements of the biometric database.

[0060] Back to Figure 2 , in step S2, when the similarity between the input biometric and the target reference biometric is greater than the first threshold, feature fusion is performed on the input biometric and the target reference biometric to obtain the fused feature.

[0061] In some embodiments of the present disclosure, a first threshold is preset and greater than a second threshold. The second threshold is the minimum similarity between the input biometric feature for identity recognition and the reference biometric feature when the identity recognition is successful. The application scenario of some embodiments of the present disclosure is set as follows: when the similarity between the input biometric feature and the target reference biometric feature is not less than the second threshold, the identity verification of the input biometric feature is allowed to pass; otherwise, the identity verification of the input biometric feature is not allowed to pass.

[0062] For example, in one embodiment, the second threshold is 95% and the first threshold is 97%. When the similarity between the input biometric feature and the target reference biometric feature is not less than 95%, the identity verification of the input biometric feature is allowed to pass; otherwise, the identity verification of the input biometric feature is not allowed to pass. When the similarity between the input biometric feature and the target reference biometric feature is greater than 97%, the biometric database is updated; otherwise, the biometric database is not updated. That is, compared with the judgment of the identity of the input biometric feature, a higher similarity requirement is proposed for the update of the biometric database, so that the accuracy of the update of the biometric database can be improved.

[0063] In deep learning, both the input biometric feature and the reference biometric feature are vectors. Feature fusion refers to a method of transforming two or more biometric feature vectors into one biometric feature vector.

[0064] In some embodiments of the present disclosure, feature fusion is performed on the input biometric feature and the target reference biometric feature to obtain a fused feature, including:

[0065] Feature fusion is performed on the input biometric feature and the target reference biometric feature according to the function feat_out = feat_cap * (1 - momentum) + feat_old * momentum, where feat_out is the fused feature, feat_cap is the input biometric feature, feat_old is the target reference biometric feature, and momentum is the momentum coefficient, and 0 ≤ momentum < 1.

[0066] In some embodiments, the momentum coefficient is greater than or equal to 0.9 and less than 1. For example, in one embodiment, the value of momentum is 0.95. In this way, the update of the target reference biometric feature is relatively gentle and subtle, making the accuracy of biometric recognition higher.

[0067] In some other embodiments of the present disclosure, feature fusion is performed on the input biometric feature and the target reference biometric feature to obtain a fused feature, including: concatenating the vector components of the input biometric feature and the target reference biometric feature to obtain the fused feature.

[0068] Back to Figure 2, in step S3, update the target reference biometric feature to the fused feature.

[0069] In the biometric database, each reference biometric feature corresponds to an identity document (ID). In some embodiments, the above step S3 includes:

[0070] Search for the target reference biometric feature in the biometric database according to the identity document of the target reference biometric feature;

[0071] Replace the target reference biometric feature with the fused feature.

[0072] As Figure 3 shown, the method for updating the biometric database in some embodiments of the present disclosure is used in the face recognition scenario. The method flow for face recognition and updating the face feature database includes the following steps S21 to step S33.

[0073] In step S21, obtain multiple pictures including face regions taken by an image acquisition device.

[0074] For example, before user Wang enters the company's office area, he needs to complete the check-in through an attendance card punching machine with face recognition function. The camera of the attendance card punching machine captures a series of video frames P1, P2, P3..., P12 including Wang's face region.

[0075] In step S22, according to the set scoring rules, perform quality scoring on the multiple pictures including face regions obtained in the above step S21, and filter out the pictures with quality scores lower than the set threshold.

[0076] For example, since the face region of video frame P11 is relatively blurred and the face region of video frame P12 is incomplete, the quality scoring fails, so video frames P11 and P12 are filtered out in this step S22.

[0077] In step S23, for each picture including a face region, use a face feature extraction algorithm to extract the face feature from the picture. For example, extract face features F1, F2, F3..., F10 from video frames P1, P2, P3..., P10 respectively.

[0078] In step S23, the face feature extraction algorithm used is not limited.

[0079] For example, in some embodiments, the face feature extraction algorithm adopts the Active Shape Model (ASM) algorithm. ASM is an algorithm based on the Point Distribution Model (PDM). In the PDM algorithm, the geometric shapes of objects with similar appearances, such as human faces, human hands, hearts, lungs, etc., can be represented by a shape vector formed by sequentially concatenating the coordinates of several key feature points. The process of extracting face features from a picture using the ASM algorithm includes: detecting the face region using a classifier; using the trained model to find a fixed number of feature points on the face (for example, 68 feature points); recording the coordinate positions of each feature point and sequentially concatenating them to form a shape vector.

[0080] For example, in some other embodiments, the face feature extraction algorithm adopts the Active Appearance Model (AAM) algorithm. The AAM algorithm further performs statistical modeling on the texture based on the ASM algorithm, and further fuses the two statistical models of shape and texture into an appearance model. After unifying the dimensions of the shape and texture features in the AAM algorithm, the modeling and search processes are basically the same as those of ASM.

[0081] In step S24, the face features F1, F2, F3..., F10 extracted in step S23 are subjected to feature fusion to obtain the input face feature F.

[0082] For example, among the foregoing series of video frames P1, P2, P3..., P10, the picture of video frame P2 is slightly blurred. If only the face feature F2 is extracted from this video frame P2 as the input face feature for face recognition, then to a certain extent, it may affect the accuracy of the recognition result. In this embodiment, feature fusion processing is performed on the face features of this series of video frames, and the influence of the image quality of video frame P2 on the input face feature can be ignored. Therefore, the accuracy of face recognition can be improved.

[0083] In step S25, the reference face features M1, M2, M3... in the face feature library are traversed, and the input face feature F is compared with each reference face feature for similarity to obtain a similarity list as shown in Table 1 below.

[0084]

[0085]

[0086] Table 1 Similarity list

[0087] In step S26, the similarity list is screened, and the similarity comparison data with a confidence level not less than the confidence level threshold is retained as the first screening result, as shown in Table 2 below.

[0088] Reference face features M1 M6 M8 Comparison similarity 98％ 99％ 93％

[0089] Table II First Screening Result List

[0090] When making an estimate of the population parameter through sampling, due to the randomness of the sample, the conclusion is uncertain. The confidence level refers to the probability that the population parameter value falls within a certain interval of the sample statistic value. In the similarity list, if the confidence level of the similarity comparison data is less than the confidence level threshold, the face recognition result is considered untrustworthy and the comparison data is excluded; if the confidence level of the similarity comparison data is not less than the confidence level threshold, the face recognition result is considered trustworthy and the comparison data is retained.

[0091] In step S27, the first screening result is screened again, and the comparison data of the reference face feature whose associated attribute is consistent with the input face feature F is retained as the second screening result, as shown in Table III below. In some embodiments, the associated attribute includes at least one of gender, age group, and physical morphology category.

[0092] Reference face features M1 M8 Comparison similarity 98％ 93％

[0093] Table III Second Screening Result List

[0094] For example, the gender corresponding to the reference face feature M6 in Table II above is inconsistent with the gender obtained based on the input face feature. Therefore, the comparison data related to the reference face feature M6 is excluded in this step S27.

[0095] In step S28, the reference face feature corresponding to the maximum similarity value in the second screening result is used as the target reference face feature M1, and the recognition result is output based on this target reference face feature M1. The recognition result includes the above maximum similarity value (98%), the face image restored based on the target reference face feature M1, and the identity identifier of Wang, such as Wang's employee number 00345.

[0096] Since the reference face features with a confidence level less than the confidence level threshold have been excluded in the previous step S27, and the reference face features with an associated attribute inconsistent with the input face feature F have been excluded in the previous step S28, the computational amount of this step S28 is greatly reduced, and the output result is also more accurate.

[0097] In step S29, the similarity of the above input biometric feature and the target reference biometric feature is obtained.

[0098] In step S30, it is judged whether the similarity (98%) of the comparison between the input biometric feature and the target reference biometric feature is greater than the first threshold (for example, the first threshold is 95%). If so, the process proceeds to step S31; otherwise, the process ends.

[0099] In step S31, the target reference face feature M1 is retrieved from the face feature database according to the identity identifier corresponding to the target reference face feature M1, such as the employee number 00345 of Wang.

[0100] In step S32, the input face feature F and the target reference face feature M1 are subjected to feature fusion to obtain the fused feature M1'.

[0101] In step S33, the target reference face feature M1 in the face feature database is updated to the fused feature M1', that is, the target reference face feature M1 in the face feature database is replaced with the fused feature M1'.

[0102] As Figure 4 shown, some embodiments of the present disclosure also provide a device 400 for updating a biometric database. The biometric database includes at least one reference biometric. The device 400 for updating the biometric database includes:

[0103] An acquisition unit 41, configured to acquire the similarity between the input biometric and the target reference biometric, where the input biometric is a biometric extracted from multimedia data, and the target reference biometric is the reference biometric with the highest identity matching degree with the input biometric among at least one reference biometric;

[0104] A fusion unit 42, configured to perform feature fusion on the input biometric and the target reference biometric to obtain a fused feature when the similarity is greater than a first threshold;

[0105] An update unit 43, configured to update the target reference biometric to the fused feature.

[0106] The above embodiments of the present disclosure can automatically update the biometric database according to program settings. Compared with the fixed reference biometrics in the related art, the reference biometrics in the embodiments of the present disclosure can be dynamically adjusted following the changes of individual biometrics. The updated biometric database is applied to biometric recognition, which can improve the recognition accuracy and thus enhance the information security.

[0107] As Figure 5 shown, some embodiments of the present disclosure also provide an electronic device 500, including: a memory 51 and a processor 52 coupled to the memory 51. The processor 52 is configured to execute the method for updating the biometric database according to any of the foregoing embodiments based on instructions stored in the memory 51.

[0108] The memory 51 may include a Random Access Memory (RAM), or may also include a Non-Volatile Memory (NVM), such as at least one disk memory. The memory 51 may also be at least one storage device located far away from the aforementioned processor 52.

[0109] The aforementioned processor 52 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be 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, discrete hardware components.

[0110] Some embodiments of the present disclosure also provide a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the method for updating a biometric database as described in any of the foregoing technical solutions.

[0111] In addition, in some other embodiments of the present disclosure, there is also provided a computer program product containing instructions, and when it runs on a computer, it causes the computer to execute the method for updating a biometric database in any of the above embodiments.

[0112] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions according to the embodiments of the present disclosure are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid state disk (SSD)).

[0113] 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 actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0114] Each embodiment in this specification is described in a related manner. The same or similar parts among the embodiments can be referred to each other, and the differences between each embodiment and other embodiments are emphasized. In particular, for the embodiments of the device, electronic device, and computer-readable storage medium, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments.

[0115] The above are only the preferred embodiments of the present disclosure and are not intended to limit the protection scope of the present disclosure. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present disclosure are all included in the protection scope of the present disclosure.

Claims

1. A method for updating a biometric database, the biometric database including at least one reference biometric, characterized in that, The method includes: Obtaining the similarity between the input biometric feature and the target reference biometric feature, where the input biometric feature is the biometric feature extracted from the multimedia data, and the target reference biometric feature is the reference biometric feature with the highest identity matching degree with the input biometric feature among the at least one reference biometric feature; When the similarity is greater than the first threshold, performing feature fusion on the input biometric feature and the target reference biometric feature to obtain the fused feature; Searching for the target reference biometric feature in the biometric feature library according to the identity identifier of the target reference biometric feature; Replacing the target reference biometric feature with the fused feature.

2. The method for updating the biometric feature library according to claim 1, wherein: The input biometric feature is obtained by performing biometric feature extraction on a single picture containing biometric features; or The input biometric feature is obtained by performing biometric feature extraction on multiple pictures containing biometric features and then performing biometric feature fusion.

3. The method for updating the biometric feature library according to claim 1, wherein: Obtaining the similarity between the input biometric feature and the target reference biometric feature includes: after completing identity recognition based on the input biometric feature, obtaining the similarity between the input biometric feature and the target reference biometric feature; The first threshold is greater than the second threshold, and the second threshold is: the minimum similarity between the input biometric feature indicating successful identity recognition and the reference biometric feature when performing identity recognition based on the input biometric feature.

4. The method for updating a biometric database according to claim 1, wherein It further includes: Before obtaining the similarity between the input biometric feature and the target reference biometric feature, determining the target reference biometric feature according to the similarity between the input biometric feature and the reference biometric feature and the result information of whether the association attributes between the input biometric feature and the reference biometric feature are consistent; Wherein, the confidence level of the similarity between the input biometric feature and the reference biometric feature is not less than the confidence level threshold, and the association attributes include at least one of gender, age group, and physical form category.

5. The method for updating a biometric database according to claim 1, wherein Performing feature fusion on the input biometric feature and the target reference biometric feature to obtain the fused feature includes: Performing feature fusion on the input biometric feature and the target reference biometric feature according to the function feat_out = feat_cap * (1 - momentum) + feat_old * momentum; Wherein, feat_out is the fused feature, feat_cap is the input biometric feature, feat_old is the target reference biometric feature, momentum is the momentum coefficient, and 0 ≤ momentum < 1.

6. The method for updating a biometric database according to claim 5, wherein 0.9 ≤ momentum < 1.

7. The method for updating a biometric database according to any one of claims 1-6, characterized in that The biometric feature library includes a face feature library, a palm feature library, a skin feature library, an auricle feature library, a gait feature library, or a voice feature library.

8. An apparatus for updating a biometric database, the biometric database including at least one reference biometric, characterized in that, The device includes: An acquisition unit, configured to acquire a similarity between an input biometric feature and a target reference biometric feature for comparison, where the input biometric feature is a biometric feature extracted from multimedia data, and the target reference biometric feature is the reference biometric feature with the highest identity matching degree with the input biometric feature among the at least one reference biometric feature; A fusion unit, configured to perform feature fusion on the input biometric feature and the target reference biometric feature to obtain a fused feature when the similarity is greater than a first threshold; An update unit, configured to search for the target reference biometric feature from the biometric feature library according to the identity identifier of the target reference biometric feature; and replace the target reference biometric feature with the fused feature.

9. An electronic device, characterized in that, It includes: a memory and a processor coupled to the memory, and the processor is configured to execute the method for updating the biometric feature library according to any one of claims 1-7 based on instructions stored in the memory.

10. A computer-readable storage medium, characterized in that, A computer program is stored thereon, and when the computer program is executed by the processor, it implements the method for updating the biometric feature library according to any one of claims 1-7.

Images