Processing method and device based on face recognition, storage medium and program product

Abstract

The invention discloses a processing method and device based on face recognition, a storage medium and a program product. The method relates to the field of artificial intelligence, and comprises the steps of obtaining N similarities output after face recognition is performed on a target object, the similarities comprising similarity parameters between a face identity obtained by performing face recognition on the target object through a face recognition system and an object identity in a historical time period, the object identity is pre-stored standard identity information associated with the target object; obtaining K intervals, and performing risk assessment on each similarity based on the interval to which each similarity belongs to obtain risk values corresponding to N similarities; accumulating the risk values corresponding to the N similarities to obtain a risk accumulation value; and when the risk accumulation value exceeds the risk limit of the target object, generating early warning information. Through the method and the device, the problem of low safety caused by risk control according to a similarity threshold value of face recognition in related technologies is solved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence, and more specifically, to a face recognition-based processing method, apparatus, storage medium, and program product. Background Technology

[0002] Risk control refers to the process of identifying and assessing potential risks and taking measures to reduce the likelihood and impact of these risks when using artificial intelligence for decision-making.

[0003] The main steps of a face recognition system in related technologies are: (1) Obtaining the face feature vector contained in the input face photo. This step can be further divided into face image acquisition, face detection, face alignment, and face feature extraction. (2) Comparing the extracted face feature vector with the face feature vector stored in the database. Specifically, the similarity (Euclidean distance or cosine similarity, etc.) between the extracted vector and the vector in the database is calculated sequentially. (3) Outputting the results. If the similarity exceeds a certain threshold, the match is successful, and the user identity corresponding to the successfully matched face feature vector in the database is output.

[0004] However, facial recognition systems based on similarity thresholds lack risk control for successful matches with low similarity scores. On the one hand, setting the threshold too high will increase the number of failed facial recognition attempts, impacting user experience; on the other hand, setting the threshold too low will reduce the difficulty for attackers, making it easier to forge facial features and pass facial recognition. Therefore, deploying facial recognition systems often requires extensive testing and adjustments to balance usability and security in real-world environments, consuming significant manpower, while still neglecting the security risks posed by successful low-similarity matches (e.g., impacting the security of financial transactions).

[0005] There is currently no effective solution to the problem of low security in risk control based on facial recognition similarity thresholds in related technologies. Summary of the Invention

[0006] The main objective of this application is to provide a face recognition-based processing method, device, storage medium, and program product to solve the problem of low security in related technologies that rely on face recognition similarity thresholds for risk control.

[0007] To achieve the above objectives, according to one aspect of this application, a face recognition-based processing method is provided. The method includes: acquiring N similarity scores output after performing face recognition on a target object, wherein the similarity scores include: similarity parameters between the facial identity obtained through face recognition of the target object during financial transactions conducted on the target object within a historical time period, and the object identity being pre-stored standard identity information associated with the target object, where N is a positive integer; acquiring K intervals, and performing a risk assessment on each similarity score based on the interval to which each similarity score belongs, obtaining N risk values ​​corresponding to the similarity scores, where K is a positive integer; accumulating the N risk values ​​corresponding to the similarity scores to obtain a cumulative risk value; and generating a warning message if the cumulative risk value exceeds the risk limit of the target object, wherein the warning message indicates that the face recognition result of the target object poses a security risk.

[0008] Further, the risk values ​​corresponding to the N similarities are accumulated to obtain a risk accumulation value, including: obtaining the generation time of each similarity; based on the generation time associated with each similarity, using a time decay strategy to generate a decay coefficient for each similarity, wherein the decay coefficient is greater than or equal to 0 and less than or equal to 1, and the earlier the generation time of each similarity, the smaller the decay coefficient corresponding to that generation time; and based on the decay coefficients of the N similarities, the risk values ​​corresponding to the N similarities are accumulated to obtain the risk accumulation value.

[0009] Further, based on the decay coefficients of the N similarities, the risk values ​​corresponding to the N similarities are accumulated to obtain the risk accumulation value, including: calculating the product of the risk value corresponding to each similarity and the decay coefficient of that similarity to obtain the target value of each similarity; and calculating the sum of the target values ​​of the N similarities to obtain the risk accumulation value.

[0010] Further, obtaining K intervals includes: obtaining a similarity threshold, wherein the similarity threshold includes: the minimum similarity required to determine if a face recognition result is acceptable; determining a target interval based on the similarity threshold, wherein the target interval includes: the range of similarity required to determine if a face recognition result is acceptable; and segmenting the target interval to obtain K intervals.

[0011] Further, based on the interval to which each similarity belongs, a risk assessment is performed on each similarity to obtain N risk values ​​corresponding to each similarity, including: determining the risk value corresponding to each of the K intervals, wherein the risk value corresponding to the interval closer to the similarity threshold is higher than the risk value corresponding to the interval farther from the similarity threshold; and determining N risk values ​​corresponding to each similarity based on the interval to which each similarity belongs and the risk value corresponding to each interval.

[0012] Furthermore, if the accumulated risk value exceeds the risk limit of the target object, the method further includes: if the target object requests to conduct financial business, performing secondary verification of the target object's identity to obtain a verification result; and if the verification result indicates that the target object's identity verification is successful, conducting the financial business.

[0013] Furthermore, before generating an early warning message when the accumulated risk value exceeds the risk limit of the target object, the process includes: determining the object type of the target object; obtaining a target mapping relationship, wherein the target mapping relationship includes the relationship between the object type and the risk limit; and determining the risk limit of the target object based on the object type and the target mapping relationship.

[0014] To achieve the above objectives, according to another aspect of this application, a face recognition-based processing apparatus is provided. The apparatus includes: a first acquisition unit, configured to acquire N similarity scores output after face recognition of a target object, wherein the similarity scores include similarity parameters between the face identity obtained by face recognition of the target object through a face recognition system during the processing of financial transactions involving the target object within a historical time period, wherein the object identity is pre-stored standard identity information associated with the target object, and N is a positive integer; an evaluation unit, configured to acquire K intervals and, based on the interval to which each similarity score belongs, perform a risk assessment on each similarity score to obtain N risk values ​​corresponding to the similarity scores, wherein K is a positive integer; an accumulation unit, configured to accumulate the risk values ​​corresponding to the N similarity scores to obtain a risk accumulation value; and a generation unit, configured to generate warning information when the risk accumulation value exceeds the risk limit of the target object, wherein the warning information indicates that the face recognition result of the target object has a security risk.

[0015] Further, the accumulation unit includes: a first acquisition subunit, used to acquire the generation time of each similarity; a generation subunit, used to generate a decay coefficient for each similarity based on the generation time associated with each similarity using a time decay strategy, wherein the decay coefficient is greater than or equal to 0 and less than or equal to 1, and the earlier the generation time of each similarity, the smaller the decay coefficient corresponding to that generation time; and an accumulation subunit, used to accumulate the risk values ​​corresponding to N similarities based on the decay coefficients of N similarities to obtain the risk accumulation value.

[0016] Further, the accumulation subunit includes: a first calculation module, used to calculate the product of the risk value corresponding to each similarity and the decay coefficient of the similarity, to obtain the target value of each similarity; and a second calculation module, used to calculate the sum of the target values ​​of N similarities, to obtain the risk accumulation value.

[0017] Furthermore, the evaluation unit includes: a second acquisition subunit, used to acquire a similarity threshold, wherein the similarity threshold includes: the minimum similarity required for a face recognition result to pass recognition; a first determination subunit, used to determine a target interval based on the similarity threshold, wherein the target interval includes: the range of similarity required for a face recognition result to pass recognition; and a segmentation subunit, used to segment the target interval to obtain K intervals.

[0018] Furthermore, the evaluation unit includes: a second determining subunit, configured to determine the risk value corresponding to each of the K intervals, wherein the risk value corresponding to the interval closer to the similarity threshold is higher than the risk value corresponding to the interval farther from the similarity threshold; and a third determining subunit, configured to determine the risk values ​​corresponding to N similarities based on the interval to which each similarity belongs and the risk value corresponding to each interval.

[0019] Furthermore, the face recognition-based processing device also includes: a verification unit, configured to perform secondary verification of the identity of the target object when the accumulated risk value exceeds the risk limit of the target object and the target object requests to conduct financial business, and obtain a verification result; and a processing unit, configured to process the financial business when the verification result indicates that the identity verification of the target object is successful.

[0020] Furthermore, the face recognition-based processing device further includes: a first determining unit, configured to determine the object type of the target object before generating a warning message when the accumulated risk value exceeds the risk limit of the target object; a second acquiring unit, configured to acquire a target mapping relationship, wherein the target mapping relationship includes the relationship between the object type and the risk limit; and a second determining unit, configured to determine the risk limit of the target object based on the object type and the target mapping relationship.

[0021] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the face recognition-based processing method.

[0022] According to another aspect of this application, an electronic device is provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes the face recognition-based processing method during runtime.

[0023] According to another aspect of this application, a computer program product is provided, including computer instructions that, when executed by a processor, implement the steps of the face recognition-based processing method.

[0024] In this application, the following method is adopted: N similarity scores are obtained after facial recognition of the target object. These similarity scores include similarity parameters between the facial identity obtained through facial recognition of the target object during financial transactions within a historical time period, and the object identity is pre-stored standard identity information associated with the target object. N is a positive integer. K intervals are obtained, and based on the interval to which each similarity score belongs, a risk assessment is performed on each similarity score to obtain N risk values ​​corresponding to these similarities. K is a positive integer. The risk values ​​corresponding to the N similarities are accumulated to obtain a cumulative risk value. If the cumulative risk value exceeds the target object's risk limit, a warning message is generated. This warning message indicates that the facial recognition result of the target object has a security risk, thus solving the technical problem of low security in related technologies that rely solely on facial recognition similarity thresholds for risk control. In this application, by accumulating the risk value of facial recognition similarity and issuing warnings based on the accumulated risk value, the risk control based solely on the facial recognition similarity threshold in related technologies is avoided, which ignores the security risks brought about by successful low-similarity matching. This achieves the technical effect of improving the security of facial recognition risk control. Attached Figure Description

[0025] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0026] Figure 1 A hardware structure block diagram of a computer terminal for implementing a face recognition-based processing method is shown.

[0027] Figure 2 This is a flowchart of a face recognition-based processing method provided according to an embodiment of this application;

[0028] Figure 3 This is a schematic diagram of a face recognition-based processing system provided according to an embodiment of this application;

[0029] Figure 4 This is a schematic diagram of a face recognition-based processing device provided according to an embodiment of this application;

[0030] Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0031] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0032] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0033] It should be noted that the face recognition-based processing method and apparatus in this application can be used for risk control of face recognition systems in the field of artificial intelligence, or for risk control of face recognition systems in any field other than artificial intelligence. This application does not limit the application field of the face recognition-based processing method and apparatus.

[0034] First, some nouns or terms that appear in the description of the embodiments of this application shall be interpreted as follows:

[0035] A face recognition system (or facial recognition system) is a pre-trained and deployed system for classification tasks. It can infer the user's identity from an input face image (i.e., a face image) and give a confidence score (i.e., similarity).

[0036] It should be noted that the information collected in this application (including but not limited to user device information, user personal information, facial information, etc.) and data (including but not limited to data used for display, data for analysis, facial data, etc.) are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of this data all comply with relevant laws, regulations, and standards, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding access points are provided for users to choose to authorize or refuse. For example, interfaces are set up between this system and relevant users or organizations, providing users with corresponding access points to choose to agree to or refuse automated decision results; if the user chooses to refuse, the process proceeds to the expert decision-making stage.

[0037] This invention can be applied to various software products, control systems, and client applications (including but not limited to mobile clients and PCs) of financial institutions in identity verification scenarios for handling financial transactions. Taking a software product as an example, the facial recognition system of the software product installed on the mobile client can improve the security of financial institutions' business operations (including but not limited to: transfers, wealth management, funds, bill payments, account inquiries, advertising, recommendations, and other business functions).

[0038] Example 1

[0039] According to an embodiment of this application, a method embodiment of a face recognition-based processing method is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0040] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware block diagram of a computer terminal (or mobile device) for implementing a face recognition-based processing method is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0041] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0042] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the face recognition-based processing method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the face recognition-based processing method described above. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0043] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0044] The display can be, for example, a touchscreen liquid crystal display (LCD), which allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0045] Under the aforementioned operating environment, this application provides the following: Figure 2 The processing method based on face recognition is shown. Figure 2 This is a flowchart of a face recognition-based processing method according to Embodiment 1 of this application.

[0046] Step S201: Obtain N similarity values ​​output after performing facial recognition on the target object. The similarity values ​​include: similarity parameters between the facial identity and the object identity obtained by performing facial recognition on the target object through the facial recognition system during the financial business transactions of the target object within a historical time period. The object identity is the pre-stored standard identity information associated with the target object, and N is a positive integer.

[0047] The target object mentioned above can be a customer of a financial institution, and the N similarities mentioned above can include: the similarity between the facial identity obtained by facial recognition of the target object and the object's identity during the financial business transactions with the target object within a historical time period.

[0048] Step S202: Obtain K intervals, and based on the interval to which each similarity belongs, perform a risk assessment on each similarity to obtain N risk values ​​corresponding to the similarities, where K is a positive integer.

[0049] The aforementioned K intervals can be K similarity intervals. In this embodiment, the K intervals can be determined based on the similarity threshold of the face recognition system. For example, if the similarity threshold of the face recognition system is 0.8, then face samples with a similarity threshold exceeding 0.8 are considered successfully recognized (i.e., verified) and correspond to a face identity. In this case, the range from 0.8 to 1.0 can be divided into K equally divided intervals, and then a different risk value can be assigned to each interval.

[0050] In this embodiment, the risk value corresponding to each similarity can be determined based on the interval to which each similarity belongs.

[0051] Step S203: Accumulate the risk values ​​corresponding to N similarities to obtain the cumulative risk value.

[0052] In this embodiment, the risk value corresponding to each of the N similarities can be accumulated to obtain the cumulative risk value.

[0053] Step S204: If the cumulative risk value exceeds the risk limit of the target object, generate a warning message, wherein the warning message is used to indicate that there is a security risk in the facial recognition result of the target object.

[0054] The risk limit for a target object can include: a preset upper limit for the cumulative risk value of the target object, used to determine whether there is a security risk in the facial recognition result of the target object. For example, if the cumulative risk value of the target object exceeds the risk limit of the target object, an early warning can be generated to indicate that there is a security risk in the facial recognition of the target object. This avoids the risk brought by facial recognition results with low similarity matching when the similarity threshold of the facial recognition system in related technologies is set too low, which reduces the difficulty of attack for attackers. In other words, it reduces the difficulty of forging facial features to pass through facial recognition.

[0055] Through the above steps, in this embodiment, by accumulating the risk value of facial recognition similarity and issuing warnings based on the accumulated risk value, the risk control method used in related technologies, which relies solely on the facial recognition similarity threshold and ignores the security risks associated with successful low-similarity matches, is avoided. This achieves the technical effect of improving the security of facial recognition risk control. Furthermore, it solves the technical problem of low security in related technologies that rely solely on facial recognition similarity thresholds for risk control.

[0056] Optionally, in the face recognition-based processing method provided in this application embodiment, the risk values ​​corresponding to N similarities are accumulated to obtain a risk accumulation value, including: obtaining the generation time of each similarity; generating a decay coefficient for each similarity based on the generation time associated with each similarity using a time decay strategy, wherein the decay coefficient is greater than or equal to 0 and less than or equal to 1, and the earlier the generation time of each similarity, the smaller the decay coefficient corresponding to that generation time; and accumulating the risk values ​​corresponding to the N similarities based on the decay coefficients of the N similarities to obtain a risk accumulation value.

[0057] The aforementioned time decay strategy can include a time decay function, which can be linear or nonlinear, provided it meets the following characteristics: the time decay function can generate a decay coefficient sequence, with the decay coefficients arranged from smallest to largest (e.g., monotonically decreasing with time), and the value range can be 0-1. In this embodiment, the input to the time decay function can be the time interval between the similarity generation time and the current time, and the output of the time decay function can be the decay coefficient of that similarity. There are N decay coefficients corresponding to similarities, which can form a time decay sequence. Multiplying the decay coefficients in the decay coefficient sequence by multiple historical risk values ​​(risk values ​​corresponding to N similarities) and then summing them yields the cumulative risk value. In this case, the decay coefficients corresponding to the risk values ​​associated with earlier similarities are close to 0, making their contribution to risk accumulation almost zero; simultaneously, the decay coefficients of the risk values ​​associated with the most recent similarities are close to 1, making their contribution to risk accumulation almost equal to their risk value.

[0058] By employing a time decay strategy to accumulate risk values—specifically, using a time decay function to discount early risk values—the risk limit can be reduced from being overdrawn due to the continuous accumulation of risk values. This enables effective management and dynamic monitoring of risks in facial recognition systems.

[0059] Optionally, in the face recognition-based processing method provided in this application embodiment, the risk values ​​corresponding to the N similarities are accumulated based on the attenuation coefficients of the N similarities to obtain a risk accumulation value, including: calculating the product of the risk value corresponding to each similarity and the attenuation coefficient of that similarity to obtain a target value for each similarity; and calculating the sum of the target values ​​of the N similarities to obtain the risk accumulation value.

[0060] In this embodiment, the product of the risk value corresponding to each similarity and the decay coefficient of that similarity can be calculated to obtain the target value of each similarity; the sum of the target values ​​of N similarities is calculated to obtain the cumulative risk value, thus achieving the purpose of accumulating the historical risk value of the target object.

[0061] Optionally, in the face recognition-based processing method provided in this application embodiment, obtaining K intervals includes: obtaining a similarity threshold, wherein the similarity threshold includes: the minimum similarity used to determine that the face recognition result is acceptable; determining a target interval based on the similarity threshold, wherein the target interval includes: the similarity range used to determine that the face recognition result is acceptable; and segmenting the target interval to obtain K intervals.

[0062] The aforementioned similarity threshold may include: the minimum similarity required to determine if a face recognition result is successful. For example, if the similarity exceeds the similarity threshold, the face recognition is successful (i.e., the facial identity information obtained by the target object through face recognition matches the standard identity information); if the similarity does not exceed the similarity threshold, the face recognition is unsuccessful (i.e., the facial identity information obtained by the target object through face recognition does not match the standard identity information).

[0063] In this embodiment, the similarity range (i.e., the target interval) of the face recognition result can be determined according to the similarity threshold. The target interval can be divided into K intervals by a target segmentation strategy. The target segmentation strategy can be equal division, or in an optional example, unequal division.

[0064] For example, if the similarity threshold of a facial recognition system is 0.8, then facial samples with a similarity exceeding 0.8 can be considered successfully recognized and correspond to a facial identity. In this case, the range of 0.8 to 1.0 can be divided into K equal intervals, and then a different risk value can be assigned to each interval. Assuming an interval is divided every 0.5, then a similarity of 0.8 to 0.85 can be assigned a higher risk value, while 0.95 to 1.0 should be assigned a risk value close to zero. This risk measurement method considers the difference in risk between low and high similarity when both are successfully recognized, which is more in line with real-world scenarios. At the same time, a high risk value for low similarity will cause risk to accumulate more quickly, while a low risk value for high similarity will slow down the rate of risk accumulation, thus ensuring a more granular balance between the usability and security of the AI ​​system.

[0065] Optionally, in the face recognition-based processing method provided in this application embodiment, a risk assessment is performed on each similarity based on the interval to which each similarity belongs, to obtain risk values ​​corresponding to N similarities, including: determining the risk value corresponding to each interval in K intervals, wherein the risk value corresponding to the interval closer to the similarity threshold is higher than the risk value corresponding to the interval farther from the similarity threshold; and determining the risk values ​​corresponding to N similarities based on the interval to which each similarity belongs and the risk value corresponding to each interval.

[0066] In this embodiment, different risk values ​​can be assigned to different intervals. The closer an interval is to the similarity threshold, the higher the corresponding risk value; the farther an interval is from the similarity threshold, the lower the corresponding risk value. For example, if the similarity threshold is 0.8, the range from 0.8 to 1.0 is divided into k equal intervals, and then a different risk is assigned to each interval. Assuming an interval is divided every 0.5, then the similarity between 0.8 and 0.85 will be assigned a larger risk value, while the range from 0.95 to 1.0 should be assigned a risk value that is almost zero. It should be noted that the range of risk values ​​in this embodiment can be flexibly set according to the actual scenario. For example, it can be set to [0, 10] or [0, 20]. The specific range needs to match the risk accumulation mechanism and risk limit to avoid a single risk value being too large, causing the limit to be quickly exhausted, or a single risk value being too small, causing the system to sluggish. Furthermore, the range of risk values ​​can be dynamically adjusted to balance security, flexibility, and maintainability.

[0067] In this embodiment, the risk value corresponding to the interval to which each similarity belongs can be determined as the risk value corresponding to that similarity, thereby improving the ease of use and security of risk control in the face recognition system.

[0068] Optionally, in the face recognition-based processing method provided in this application embodiment, if the cumulative risk value exceeds the risk limit of the target object, the method further includes: if the target object requests to conduct financial business, performing secondary verification of the target object's identity to obtain a verification result; and if the verification result indicates that the target object's identity verification is successful, conducting financial business.

[0069] In this embodiment, the cumulative risk value accumulated by a target object through all identity authentication operations performed by the target object via facial recognition within a certain period of time can be continuously tracked and calculated. If the cumulative risk value exceeds the risk threshold preset for the target object, it indicates that the target object's recent authentication behavior may involve a high risk, requiring further verification and monitoring.

[0070] For example, when the accumulated risk value exceeds the risk limit, a secondary identity verification for the target can be automatically triggered. This is because, although the user's facial recognition result may indicate that their identity is correct, the accumulated risk value exceeding the limit suggests a potential security threat, such as identity theft or forged facial recognition. Secondary verification can employ various methods, including but not limited to: sending a one-time verification code (OTP) to the target's registered mobile phone or email address, requesting the target to provide other forms of identity verification (such as scanned copies of an ID card or passport), conducting real-time video identity verification, and biometric identification, to confirm the user's true identity through additional verification methods and reduce security risks.

[0071] After secondary verification, if the verification result shows that the identity is indeed genuine (i.e., the secondary verification is passed), the system can continue to process the financial business requested by the user. This ensures that even under conditions of high accumulated risk, genuine and legitimate users can still conduct their financial activities smoothly, thereby enhancing security while achieving a balance between security and user experience.

[0072] Optionally, in the face recognition-based processing method provided in this application embodiment, before generating warning information when the cumulative risk value exceeds the risk limit of the target object, the method includes: determining the object type of the target object; obtaining the target mapping relationship, wherein the target mapping relationship includes: the relationship between the object type and the risk limit; and determining the risk limit of the target object based on the object type and the target mapping relationship.

[0073] In this embodiment, target objects can be categorized into different object types based on factors such as behavioral patterns, usage frequency, and account attributes. For example, target objects can be divided into "low-risk objects," "medium-risk objects," and "high-risk objects." More specifically, they can be classified based on information such as age, occupation, and usage history. For instance, some elderly people have a low tolerance for risk and can be assigned a low-risk credit limit. This low-risk limit makes it easier for the accumulated risk value to exceed the limit (similar to credit card overdraft). In this case, the notification module can be triggered more promptly to issue warnings to administrators and users, prompting manual intervention.

[0074] The aforementioned target mapping relationship is a preset rule that defines the correspondence between different object types and their respective risk limits. This mapping relationship can be static, meaning it is determined during system deployment, or dynamic, meaning it can be adjusted according to the actual operation of the system and risk management strategies. Through this mapping relationship, the system can quickly query the risk limit corresponding to a specific object type.

[0075] The above steps enable personalized risk limit management. Risk limits can be allocated individually based on a user's risk tolerance, or default risk limits can be assigned to different groups. This design enhances the flexibility of risk control in facial recognition systems.

[0076] In this embodiment, the binary judgment method used in related technologies, which passes if the similarity is higher than the threshold and fails if it is lower than the threshold, is avoided. This ignores the security risks caused by low similarity passing authentication. By accumulating risk values, and issuing an alarm in a timely manner when the accumulated risk value exceeds the risk limit, the technical effect of improving the reliability of risk control in the face recognition system is achieved.

[0077] A hierarchical risk measurement and flexible risk accumulation method are adopted to reduce the problem of excessively rapid risk accumulation caused by direct risk measurement and accumulation. In this embodiment, risk values ​​with different similarity levels and risk values ​​from different historical periods are treated differently, so that the measurement and accumulation of risk are effectively controlled, while also paying special attention to recent cases where low similarity has passed authentication.

[0078] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0079] Example 2

[0080] Embodiment 2 of the present invention provides an optional face recognition-based processing system, which can be used to execute the face recognition-based processing method provided in Embodiment 1 of the present invention.

[0081] Figure 3 This is a schematic diagram of a face recognition-based processing system provided according to an embodiment of this application, such as... Figure 3 As shown, the processing system may include: a risk measurement module, a risk accumulation module, a risk limit management module, a configuration module, and a notification module.

[0082] Specifically, the risk measurement module can be used to calculate the risk of facial identity and similarity output by the facial recognition system (corresponding to the facial recognition system in Example 1). This calculation is typically hierarchical, and it satisfies the characteristic that the closer the similarity is to the threshold, the greater the risk. For example, if the similarity threshold of the facial recognition system is 0.8, then facial samples with a similarity exceeding 0.8 can be confirmed as successfully recognized and correspond to a facial identity. In this case, the risk measurement module can divide the range from 0.8 to 1.0 into k equally divided intervals, and then assign a different risk value to each interval. Assuming an interval is divided every 0.5, then a similarity between 0.8 and 0.85 will be assigned a larger risk value, while a similarity between 0.95 and 1.0 should be assigned a risk value close to 0. This risk measurement method considers the risk difference between low and high similarity even when both are successfully recognized, making it more consistent with real-world scenarios. Simultaneously, a large risk value for low similarity will cause it to accumulate risk faster, while a small risk value for high similarity will slow down the risk accumulation rate, thus ensuring a more granular balance between the usability and security of the artificial intelligence system.

[0083] The risk accumulation module can flexibly accumulate historical risk values ​​for the same facial identity. Specifically, it uses a time decay function to discount earlier risk values, reducing the risk overdraft caused by the continuous accumulation of risk values. The time decay function can be linear or non-linear, as long as it meets the following characteristics: the time decay function can generate a decay coefficient sequence, arranged from smallest to largest, and the value range is 0-1. Then, multiplying the decay coefficient sequence by multiple historical risk values ​​and summing them yields the accumulated risk value. In this case, the decay coefficients of earlier risk values ​​are close to 0, so their contribution to risk accumulation is almost zero; meanwhile, the decay coefficient of the most recent risk value is close to 1, so its contribution to risk accumulation is almost equal to its actual risk value.

[0084] The risk limit management module can employ a personalized risk quota management approach, assigning different risk limits to users based on their risk tolerance. For example, some elderly individuals have a low risk tolerance, so they are assigned a low-risk limit. This low-risk limit makes it easier for the accumulated risk value to exceed the limit (similar to credit card overdraft). In such cases, the notification module will be triggered more promptly, issuing warnings to administrators and users to prompt manual intervention.

[0085] The notification module can send notifications to users and administrators based on the over-limit judgment result (whether the limit has been exceeded) from the risk limit management module.

[0086] Finally, the configuration module is responsible for configuring strategies such as how the risk measurement module stratifies risks, how the risk accumulation module accumulates risks in a decaying manner, and how the risk limit management module configures personalized risk limits.

[0087] The triggering and alarm functions of the face recognition-based processing system provided in this embodiment do not affect the normal operation of the artificial intelligence system (e.g., the face recognition system), while providing security enhancements in terms of risk control.

[0088] It should be noted that the face recognition-based processing system provided in this embodiment is an add-on solution that enhances the security of face recognition systems with lower security levels. Specifically, for face recognition systems with lower thresholds, the risk measurement and accumulation provided can improve overall security. Meanwhile, for scenarios with high security requirements, higher thresholds can still be set for the face recognition system. The add-on operation of the face recognition-based processing system in this embodiment will not affect the current security of the face recognition system.

[0089] Example 3

[0090] This application also provides a face recognition-based processing device. It should be noted that the face recognition-based processing device of this application can be used to execute the face recognition-based processing method provided in this application. The face recognition-based processing device provided in this application will be described below.

[0091] According to an embodiment of this application, an apparatus for implementing the above-described face recognition-based processing method is also provided, such as... Figure 4 As shown, the device includes: a first acquisition unit 41, an evaluation unit 42, an accumulation unit 43, and a generation unit 44.

[0092] The first acquisition unit 41 is used to acquire N similarity values ​​output after performing facial recognition on the target object. The similarity values ​​include: similarity parameters between the facial identity and the object identity obtained by performing facial recognition on the target object through the facial recognition system during the financial business transactions of the target object within a historical time period. The object identity is the pre-stored standard identity information associated with the target object, and N is a positive integer.

[0093] Evaluation unit 42 is used to obtain K intervals and perform risk assessment on each similarity based on the interval to which each similarity belongs, to obtain N risk values ​​corresponding to similarities, where K is a positive integer;

[0094] Accumulation unit 43 is used to accumulate the risk values ​​corresponding to N similarities to obtain the risk accumulation value;

[0095] The generation unit 44 is used to generate early warning information when the risk accumulation value exceeds the risk limit of the target object. The early warning information is used to indicate that there is a security risk in the facial recognition result of the target object.

[0096] In the face recognition-based processing device provided in this application embodiment, the first acquisition unit 41 can acquire N similarity scores output after face recognition of the target object. The similarity scores include similarity parameters between the face identity obtained by face recognition of the target object during the financial transaction process within a historical time period, and the object identity is pre-stored standard identity information associated with the target object. N is a positive integer. The evaluation unit 42 acquires K intervals and performs a risk assessment on each similarity score based on the interval to which each similarity score belongs, obtaining N risk values ​​corresponding to the similarity scores. K is a positive integer. The accumulation unit 43 accumulates the risk values ​​corresponding to the N similarity scores to obtain a cumulative risk value. The generation unit 44 generates a warning message when the cumulative risk value exceeds the risk limit of the target object. The warning message indicates that there is a security risk in the face recognition result of the target object. This solves the technical problem of low security in related technologies that rely solely on face recognition similarity thresholds for risk control. In this embodiment, by accumulating the risk value of facial recognition similarity and issuing warnings based on the accumulated risk value, the risk control in related technologies, which only rely on the facial recognition similarity threshold and ignore the security risks brought about by successful low-similarity matching, is avoided. This achieves the technical effect of improving the security of facial recognition risk control.

[0097] Optionally, in the face recognition-based processing apparatus provided in this application embodiment, the accumulation unit includes: a first acquisition subunit, used to acquire the generation time of each similarity; a generation subunit, used to generate a decay coefficient of each similarity based on the generation time associated with each similarity using a time decay strategy, wherein the decay coefficient is greater than or equal to 0 and less than or equal to 1, and the earlier the generation time of each similarity, the smaller the decay coefficient corresponding to that generation time; and an accumulation subunit, used to accumulate the risk values ​​corresponding to N similarities based on the decay coefficients of N similarities to obtain a risk accumulation value.

[0098] Optionally, in the face recognition-based processing device provided in this application embodiment, the accumulation subunit includes: a first calculation module, used to calculate the product of the risk value corresponding to each similarity and the decay coefficient of the similarity to obtain the target value of each similarity; and a second calculation module, used to calculate the sum of the target values ​​of N similarities to obtain the risk accumulation value.

[0099] Optionally, in the face recognition-based processing apparatus provided in this application embodiment, the evaluation unit includes: a second acquisition subunit, used to acquire a similarity threshold, wherein the similarity threshold includes: the minimum similarity required to determine if the face recognition result is acceptable; a first determination subunit, used to determine a target interval based on the similarity threshold, wherein the target interval includes: the range of similarity required to determine if the face recognition result is acceptable; and a segmentation subunit, used to segment the target interval to obtain K intervals.

[0100] Optionally, in the face recognition-based processing device provided in the embodiments of this application, the evaluation unit includes: a second determining subunit, used to determine the risk value corresponding to each of the K intervals, wherein the risk value corresponding to the interval closer to the similarity threshold is higher than the risk value corresponding to the interval farther from the similarity threshold; and a third determining subunit, used to determine the risk values ​​corresponding to N similarities based on the interval to which each similarity belongs and the risk value corresponding to each interval.

[0101] Optionally, in the face recognition-based processing device provided in the embodiments of this application, the face recognition-based processing device further includes: a verification unit, used to perform secondary verification of the identity of the target object when the risk accumulation value exceeds the risk limit of the target object and the target object requests to conduct financial business, and obtain a verification result; and a processing unit, used to process financial business when the verification result indicates that the identity verification of the target object has passed.

[0102] Optionally, in the face recognition-based processing device provided in the embodiments of this application, the face recognition-based processing device further includes: a first determining unit, used to determine the object type of the target object before generating warning information when the risk accumulation value exceeds the risk limit of the target object; a second obtaining unit, used to obtain the target mapping relationship, wherein the target mapping relationship includes the relationship between the object type and the risk limit; and a second determining unit, used to determine the risk limit of the target object based on the object type and the target mapping relationship.

[0103] It should be noted that the first acquisition unit 41, evaluation unit 42, accumulation unit 43, and generation unit 44 mentioned above correspond to steps S201 to S204 in Embodiment 1. Each unit and the corresponding step implement the same instance and application scenario, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above modules can also be part of a device and run in the computer terminal 10 provided in Embodiment 1.

[0104] Example 4

[0105] Embodiments of this application may provide an electronic device. Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 5 As shown, the electronic device may include: one or more ( Figure 5 (Only one is shown) processor 502, memory 504, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0106] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the above-described methods. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0107] The processor can access information and applications stored in the memory via a transmission device to execute the following steps: Obtain N similarity scores output after facial recognition of the target object, where the similarity scores include similarity parameters between the facial identity obtained from facial recognition of the target object during financial transactions within a historical time period and the object's identity, where the object's identity is pre-stored standard identity information associated with the target object, and N is a positive integer; Obtain K intervals, and based on the interval to which each similarity belongs, perform a risk assessment on each similarity to obtain N risk values ​​corresponding to the similarity scores, where K is a positive integer; Accumulate the risk values ​​corresponding to the N similarities to obtain a cumulative risk value; If the cumulative risk value exceeds the target object's risk limit, generate a warning message, where the warning message indicates that the facial recognition result of the target object poses a security risk.

[0108] The processor can also access information and applications stored in the memory via a transmission device to perform the following steps: accumulating risk values ​​corresponding to N similarities to obtain a cumulative risk value, including: obtaining the generation time of each similarity; generating a decay coefficient for each similarity based on the generation time associated with each similarity using a time decay strategy, wherein the decay coefficient is greater than or equal to 0 and less than or equal to 1, and the earlier the generation time of each similarity, the smaller the decay coefficient corresponding to that generation time; accumulating the risk values ​​corresponding to the N similarities based on the decay coefficients of the N similarities to obtain a cumulative risk value.

[0109] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: based on the decay coefficients of N similarities, accumulate the risk values ​​corresponding to the N similarities to obtain the risk accumulation value, including: calculating the product of the risk value corresponding to each similarity and the decay coefficient of that similarity to obtain the target value of each similarity; calculating the sum of the target values ​​of the N similarities to obtain the risk accumulation value.

[0110] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: obtaining K intervals, including: obtaining a similarity threshold, wherein the similarity threshold includes: the minimum similarity used to determine that the face recognition result is recognized and passes; determining a target interval based on the similarity threshold, wherein the target interval includes: the similarity range used to determine that the face recognition result is recognized and passes; and dividing the target interval to obtain K intervals.

[0111] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: based on the interval to which each similarity belongs, perform a risk assessment on each similarity to obtain the risk values ​​corresponding to N similarities, including: determining the risk value corresponding to each interval in K intervals, wherein the risk value corresponding to the interval closer to the similarity threshold is higher than the risk value corresponding to the interval farther from the similarity threshold; and determining the risk values ​​corresponding to N similarities based on the interval to which each similarity belongs and the risk value corresponding to each interval.

[0112] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: if the cumulative risk value exceeds the risk limit of the target object, the method further includes: if the target object requests to conduct financial business, performing secondary verification of the target object's identity to obtain a verification result; if the verification result indicates that the target object's identity verification is successful, conducting financial business.

[0113] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: before generating a warning message when the cumulative risk value exceeds the risk limit of the target object, including: determining the object type of the target object; obtaining the target mapping relationship, wherein the target mapping relationship includes: the relationship between the object type and the risk limit; and determining the risk limit of the target object based on the object type and the target mapping relationship.

[0114] By employing the embodiments of this application, the risk value of facial recognition similarity is accumulated, and an early warning is issued based on the accumulated risk value. This avoids the situation in related technologies where risk control is based solely on the facial recognition similarity threshold, ignoring the security risks brought about by successful low-similarity matching. Thus, the technical effect of improving the security of facial recognition risk control is achieved.

[0115] Those skilled in the art will understand that Figure 5 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 5 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 5 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 5 The different configurations shown.

[0116] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0117] Example 5

[0118] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the face recognition-based processing method provided in Embodiment 1.

[0119] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0120] This application also provides a computer program product that, when executed on a data processing device, is adapted to perform processing method steps based on face recognition.

[0121] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0122] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0123] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0124] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0125] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0126] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0127] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A face recognition-based processing method, characterized in that, include: Obtain N similarity scores output after performing facial recognition on the target object. The similarity scores include: similarity parameters between the facial identity and the object identity obtained by performing facial recognition on the target object through the facial recognition system during the financial business transactions of the target object within a historical time period. The object identity is the standard identity information associated with the target object that is stored in advance, and N is a positive integer. Obtain K intervals, and based on the interval to which each similarity belongs, perform a risk assessment on each similarity to obtain N risk values ​​corresponding to the similarities, where K is a positive integer; Accumulate the risk values ​​corresponding to the N similarities to obtain the cumulative risk value; If the accumulated risk value exceeds the risk limit of the target object, an early warning message is generated, wherein the early warning message is used to indicate that there is a security risk in the facial recognition result of the target object.

2. The processing method according to claim 1, characterized in that, The risk values ​​corresponding to the N similarities are accumulated to obtain a cumulative risk value, including: Obtain the generation time of each of the aforementioned similarities; Based on the generation time of each similarity association, a decay coefficient for each similarity is generated using a time decay strategy, wherein the decay coefficient is greater than or equal to 0 and less than or equal to 1, and the earlier the generation time of each similarity, the smaller the decay coefficient corresponding to that generation time; Based on the decay coefficients of the N similarities, the risk values ​​corresponding to the N similarities are accumulated to obtain the cumulative risk value.

3. The processing method according to claim 2, characterized in that, Based on the decay coefficients of N similarities, the risk values ​​corresponding to the N similarities are accumulated to obtain the cumulative risk value, including: The target value for each similarity is obtained by multiplying the risk value corresponding to each similarity with the decay coefficient of that similarity. The cumulative risk value is obtained by summing the target values ​​of the N similarities.

4. The processing method according to claim 1, characterized in that, Obtain K intervals, including: Obtain a similarity threshold, wherein the similarity threshold includes: the minimum similarity required to determine if a face recognition result is successful; Based on the similarity threshold, a target interval is determined, wherein the target interval includes: a similarity range for determining the face recognition result as passed; The target interval is divided into K intervals.

5. The processing method according to claim 4, characterized in that, Based on the interval to which each similarity belongs, a risk assessment is performed on each similarity to obtain N risk values ​​corresponding to each similarity, including: Determine the risk value corresponding to each of the K intervals, wherein the risk value corresponding to the interval closer to the similarity threshold is higher than the risk value corresponding to the interval farther from the similarity threshold; Based on the interval to which each similarity belongs and the risk value corresponding to each interval, determine the risk values ​​corresponding to N similarities.

6. The processing method according to claim 1, characterized in that, If the accumulated risk value exceeds the risk limit of the target object, the method further includes: When the target object requests to conduct financial business, the identity of the target object is verified a second time to obtain the verification result; If the verification result indicates that the target object's identity verification is successful, the financial transaction will be processed.

7. The processing method according to claim 1, characterized in that, Before generating a warning message when the accumulated risk value exceeds the risk limit of the target object, the following steps are included: Determine the object type of the target object; Obtain the target mapping relationship, wherein the target mapping relationship includes: the relationship between object type and risk limit; Based on the object type and the target mapping relationship, the risk limit of the target object is determined.

8. A processing device based on face recognition, characterized in that, include: The first acquisition unit is used to acquire N similarity values ​​output after performing facial recognition on the target object. The similarity values ​​include: similarity parameters between the facial identity and the object identity obtained by performing facial recognition on the target object through the facial recognition system during the financial business transactions of the target object within a historical time period. The object identity is pre-stored standard identity information associated with the target object, and N is a positive integer. An evaluation unit is used to obtain K intervals and, based on the interval to which each similarity belongs, to perform a risk assessment on each similarity, thereby obtaining N risk values ​​corresponding to each similarity, where K is a positive integer; An accumulation unit is used to accumulate the risk values ​​corresponding to the N similarities to obtain a cumulative risk value; The generation unit is configured to generate warning information when the cumulative risk value exceeds the risk limit of the target object, wherein the warning information is used to indicate that there is a security risk in the facial recognition result of the target object.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the computer-readable storage medium is located to perform the face recognition-based processing method according to any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the face recognition-based processing method according to any one of claims 1 to 7.

Images