User verification method and device, program product and storage medium

Abstract

The embodiment of the invention discloses a user verification method and device, a program product and a storage medium, relates to the field of artificial intelligence, is also applicable to the field of financial science and technology, and comprises the following steps: obtaining handwritten image data of a to-be-verified user; the handwritten image data comprises a to-be-verified user identifier, handwriting platform information, handwriting process parameters and a handwritten image; performing content verification on the handwritten image based on the to-be-verified user identifier to obtain a content verification result of the handwritten image; the content verification result is passing or not passing; when the content verification result is successful, determining a target handwriting verification model based on the to-be-verified user identifier, and inputting the handwritten image data into the target handwriting verification model to obtain verification data output by the target handwriting verification model; and determining a handwriting verification result based on the verification data, and displaying the handwriting verification result to the to-be-verified user. According to the invention, the risk that the biological characteristic information of the user is leaked can be reduced, and the safety and accuracy of user identity verification are improved.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of artificial intelligence and can also be applied to the field of fintech, and in particular, to a user verification method, device, program product, and storage medium. Background Art

[0002] In daily fund transactions such as deposits and loans, the security of the entire transaction is very important. If there is misidentification, customers will suffer losses. Therefore, in the financial field, user verification is a key link to ensure transaction security, prevent fraud, and protect customer privacy.

[0003] Currently, the main method of user identity recognition is to determine user information through the user's biometric information such as face recognition and fingerprint recognition, so as to ensure the security of user information. Although fingerprint recognition and face recognition have a certain degree of uniqueness, these information are easy to leave traces in daily life and thus be copied or stolen, with relatively high security risks. Summary of the Invention

[0004] Embodiments of the present invention provide a user verification method, device, program product, and storage medium, which can accurately and uniquely verify users, and at the same time ensure that the user's biometric information will not be leaked, improving the accuracy and security of user verification.

[0005] In a first aspect, embodiments of the present invention provide a user verification method, including:

[0006] Obtain the handwritten image data of the user to be verified; the handwritten image data includes the user identification to be verified, handwritten platform information, handwritten process parameters, and handwritten images;

[0007] Based on the user identification to be verified, perform content verification on the handwritten image to obtain the content verification result of the handwritten image; the content verification result is either passed or not passed;

[0008] When the content verification result is passed, determine a target handwriting verification model based on the user identification to be verified, and input the handwritten image data into the target handwriting verification model to obtain the verification data output by the target handwriting verification model;

[0009] Based on the verification data, determine the handwriting verification result and display the handwriting verification result to the user to be verified.

[0010] In a second aspect, embodiments of the present invention provide a user verification device, the device includes:

[0011] A data acquisition module for acquiring the handwritten image data of the user to be verified; the handwritten image data includes the user identification to be verified, handwritten platform information, handwritten process parameters, and handwritten images.

[0012] A first verification module for performing content verification on the handwritten image based on the user identification to be verified, and obtaining a content verification result of the handwritten image; the content verification result is either passed or not passed.

[0013] A second verification module for, when the content verification result is passed, determining a target handwriting verification model based on the user identification to be verified, and inputting the handwritten image data into the target handwriting verification model to obtain verification data output by the target handwriting verification model.

[0014] A result display module for determining a handwriting verification result based on the verification data, and displaying the handwriting verification result to the user to be verified.

[0015] In a third aspect, an embodiment of the present invention further provides an electronic device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the user verification method as described in any one of the embodiments of the present invention.

[0016] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the user verification method as described in any one of the embodiments of the present invention.

[0017] In a fifth aspect, an embodiment of the present invention provides a computer program product, including a computer program, which when executed by a processor implements the user verification method as described in any one of the embodiments of the present invention.

[0018] In an embodiment of the present invention, handwritten image data of a user to be verified is obtained; the handwritten image data includes an identifier of the user to be verified, handwritten platform information, handwritten process parameters, and a handwritten image; the content of the handwritten image is verified based on the identifier of the user to be verified to obtain a content verification result of the handwritten image; the content verification result is either "passed" or "not passed"; when the content verification result is "passed", a target handwriting verification model is determined based on the identifier of the user to be verified, and the handwritten image data is input into the target handwriting verification model to obtain verification data output by the target handwriting verification model; a handwriting verification result is determined based on the verification data, and the handwriting verification result is presented to the user to be verified. The method of the embodiment of the present invention ensures a more comprehensive and accurate verification process through a dual mechanism of content verification and handwriting verification. Content verification ensures that the text content in the handwritten image is consistent with the target content, preventing verification failures caused by content errors or tampering. The target handwriting verification model corresponding to the user to be verified can accurately verify the uniqueness and authenticity of the user's handwriting, ensuring the accuracy of the user verification result. The embodiment of the present invention does not require the user's biometric information, which can reduce the risk of leakage of the user's biometric information and improve the security of user identity verification. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0020] Figure 1 The first flowchart of a user verification method provided by an embodiment of the present invention;

[0021] Figure 2 The second flowchart of a user verification method provided by an embodiment of the present invention;

[0022] Figure 3 The flowchart of a method for determining a target handwriting verification model provided by an embodiment of the present invention;

[0023] Figure 4 The structural schematic diagram of a user verification device provided by an embodiment of the present invention;

[0024] Figure 5 The structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. In addition, it should be noted that for the sake of description, only parts related to the present invention rather than all structures are shown in the accompanying drawings.

[0026] Figure 1 FIG. 1 is a first flowchart of a user authentication method provided by an embodiment of the present invention. The method of the embodiment of the present invention can accurately and uniquely authenticate a user, while ensuring that the user's biometric information is not leaked, improving the accuracy and security of user authentication. The information collected in the method of the embodiment of the present invention is information and data authorized by the user or fully authorized by all parties, and the processing of collection, storage, use, processing, transmission, provision, disclosure, and application of relevant data all comply with relevant laws, regulations, and standards of relevant countries and regions, take necessary confidentiality measures, do not violate public order and good customs, and provide corresponding operation entrances for users to choose to authorize or reject. This method can be executed by a user authentication device provided by an embodiment of the present invention, and the device can be implemented in a software and / or hardware manner. The following embodiments will be described by taking the device integrated in an electronic device as an example. The electronic device can be a server or a computer device, etc. Refer to Figure 1 , and the method specifically may include the following steps:

[0027] Step 101: Obtain the handwritten image data of the user to be authenticated.

[0028] Among them, the handwritten image data includes the user identification to be authenticated, handwritten platform information, handwritten process parameters, and handwritten images. One user corresponds to one user identification, and the user identification to be authenticated is used to uniquely mark the user to be authenticated. The handwritten platform information includes information about the device used by the user to be authenticated when handwriting, such as a graphics tablet, the handwriting screen of a smart device, the mouse of a computer, or a paper document, etc. The handwritten process parameters include the strength, stroke order, and writing speed of the user to be authenticated during the handwriting process. The handwritten image is an image containing the handwritten content of the user to be authenticated.

[0029] Specifically, when it is necessary to authenticate a user to be verified, the server can send a prompt message for inputting handwritten content and a display interface to the user to be verified. For example, when it is necessary to verify the identity of the user to be verified online in real time, the server can send a prompt message for handwritten input of corresponding content in real time to the user to be verified. When it is necessary to identify whether the signature on the paper document of the user to be verified is the user to be verified himself / herself, the server can send a prompt message for uploading a handwritten image to the user to be verified. The user to be verified can input handwritten content according to the prompt message and the display interface sent by the server, or upload a handwritten image corresponding to the paper document including the handwritten content. In an optional implementation manner, when the user to be verified inputs handwritten content through an intelligent handwriting device, such as a handwriting tablet or a screen that can be handwritten on, the server can obtain handwritten process parameters through the intelligent handwriting device. For example, the server can obtain the handwritten pressure of the user to be verified, the stroke order and speed of the user to be verified during the handwriting process, etc. through a pressure sensor of the intelligent handwriting device. When the user to be verified simulates writing on the screen with a mouse, the server can record the movement track and click events of the mouse. After the user to be verified finishes handwriting or uploads the handwritten image, the intelligent device used by the user to be verified can package and upload the handwritten image, the user identification of the user to be verified, the handwritten platform information, the handwritten process parameters, etc. to the server.

[0030] Step 102: Perform content verification on the handwritten image based on the user identification of the user to be verified to obtain a content verification result of the handwritten image.

[0031] Among them, the content verification result is either passed or not passed. Specifically, different users to be verified correspond to different handwritten contents. When the user to be verified logs in to the verification system for the first time or registers in the verification system, the server can generate a unique handwritten content corresponding to the user to be verified. When it is necessary to authenticate the user to be verified, the server can first perform content verification on the handwritten image of the user to be verified. If the content verification result is not passed, there is no need to perform handwriting verification on it anymore. If the content verification result is passed, further handwriting verification is performed on it. In this solution, performing content verification on the handwritten image based on the user identification of the user to be verified to obtain a content verification result of the handwritten image includes the following steps A1 - A2:

[0032] Step A1: Determine the target handwritten content corresponding to the user to be verified in a pre-determined user database based on the user identification of the user to be verified.

[0033] Among them, the target handwritten content is the unique handwritten content corresponding to the user to be verified generated by the server. The user database includes the unique handwritten content corresponding to each user. After obtaining the handwritten image data of the user to be verified, according to the identifier of the user to be verified, the server can find the user identifier that is the same as the identifier of the user to be verified in the user database, and determine the unique handwritten content corresponding to the user identifier as the target handwritten content.

[0034] Step A2: Extract the image content of the handwritten image through a pre-set image content extraction algorithm; perform a similarity match between the image content and the target handwritten content to obtain the content verification result of the handwritten image.

[0035] Among them, the image content extraction algorithm is pre-determined according to big data in the field, etc., and is an algorithm for extracting the image content of a handwritten image. The image content extraction algorithm can be an optical character recognition algorithm or an image extraction algorithm based on a deep learning model, etc. Specifically, after obtaining the target handwritten content, the server can preprocess the handwritten image, including operations such as denoising, binarization, and normalization, to improve the efficiency and accuracy of subsequent processing. Use the image content extraction algorithm to extract the image content from the handwritten image.

[0036] Perform a similarity match between the extracted image content and the target handwritten content. For example, the edit distance algorithm or the cosine similarity algorithm, etc., can be used to calculate the similarity between the image content and the target handwritten content. According to the similarity calculation result, determine the content verification result of the handwritten image. If the similarity is higher than a pre-set similarity threshold (such as 0.8 or 0.9), then determine that the content verification result of the handwritten image is passed. If the similarity is not higher than the similarity threshold, then determine that the content verification result of the handwritten image is not passed. By first verifying the content of the handwriting image, obviously non-compliant inputs can be quickly screened out, reducing the computing burden on the server, especially improving the verification efficiency in high-traffic financial transaction scenarios.

[0037] Step 103: When the content verification result is passed, determine the target handwriting verification model based on the identifier of the user to be verified, and input the handwritten image data into the target handwriting verification model to obtain the verification data output by the target handwriting verification model.

[0038] Among them, the target handwriting verification model is a pre-trained model for handwriting recognition of the user to be verified. In this solution, one user can correspond to one handwriting verification model. The basic structures of each handwriting verification model are the same, but the training data and model parameters are personalized for each user. The handwriting verification models corresponding to each user can adapt to the writing habits and characteristics of the handwriting devices of each user. Specifically, after determining that the content verification result is passed, the server needs to further verify the handwriting of the user to be verified. The server can retrieve the target handwriting verification model corresponding to the user to be verified in the model library according to the identifier of the user to be verified. After determining the target handwriting verification model, the server inputs the handwriting image, handwriting platform information, handwriting process parameters, etc. into the target handwriting verification model. Through the target handwriting verification model, features corresponding to the handwriting device are extracted according to the information of the handwriting device used by the user to be verified. Exemplarily, if the handwriting device used by the user to be verified is a smart handwriting tablet, the extracted features include dynamic features such as stroke order, pressure, and speed. If the handwriting device used by the user to be verified is a mouse, the extracted features include features such as trajectory speed, backward trajectory, and dwell time. If the user to be verified directly uploads a handwriting image of a paper document, the extracted features include static features such as the shape and texture of the handwriting content in the image. After extracting the features corresponding to the handwriting device, the target handwriting verification model can dynamically adjust the weights of the extracted features according to the device type and the writing habits of the user to be verified to generate weighted features. The handwriting of the user to be verified is compared and verified with the historical handwriting of the user to be verified according to the weighted features, the similarity between the handwriting of the user to be verified and the historical handwriting of the user to be verified is obtained, and verification data is output according to the similarity and the handwriting image.

[0039] In an alternative embodiment, when the content verification result is passed, the characteristic index of the handwriting image is determined based on the handwriting platform information, and the handwriting image is feature-extracted according to the characteristic index and the handwriting process parameters to obtain candidate features of the handwriting image; the candidate features are weighted and fused by the federated learning unit to obtain fused features corresponding to the candidate features; the feature map of the fused features is compressed to obtain a global feature vector; the global feature vector is dynamically weighted and adjusted based on the fully-connected layer of the federated learning unit to obtain the weighted features of the handwriting image. The similarity between the weighted features and the pre-determined historical handwriting features is calculated by the comparison unit; the verification data is obtained based on the similarity and the pre-set similarity threshold.

[0040] Step 104: Determine the handwriting verification result based on the verification data, and display the handwriting verification result to the user to be verified.

[0041] Among them, the handwriting verification result includes the identity verification result of the user to be verified and the handwritten image, etc. Specifically, after obtaining the verification data, the handwriting verification result is determined according to the verification data, and the handwriting verification result is displayed to the user to be verified. Exemplarily, if the handwriting verification result is passed, a prompt message of "Passed, please proceed to the next step" can be displayed to the user. If the handwriting verification result is not passed, a message of "Verification failed" can be displayed to the user to be verified, and the user to be verified is prompted to rewrite or contact the staff.

[0042] The technical solution of this embodiment is to obtain the handwritten image data of the user to be verified; the handwritten image data includes the user identification to be verified, the handwriting platform information, the handwriting process parameters, and the handwritten image; the content of the handwritten image is verified based on the user identification to be verified to obtain the content verification result of the handwritten image; the content verification result is passed or not passed; when the content verification result is passed, a target handwriting verification model is determined based on the user identification to be verified, and the handwritten image data is input into the target handwriting verification model to obtain the verification data output by the target handwriting verification model; the handwriting verification result is determined based on the verification data, and the handwriting verification result is displayed to the user to be verified. The technical solution of this embodiment ensures a more comprehensive and accurate verification process through a dual mechanism of content verification and handwriting verification. Content verification ensures that the text content in the handwritten image is consistent with the target content, preventing verification failures caused by content errors or tampering. The target handwriting verification model corresponding to the user to be verified can accurately verify the uniqueness and authenticity of the user's handwriting, ensuring the accuracy of the user's verification result. The technical solution of this embodiment does not require the user's biometric information, which can reduce the risk of leakage of the user's biometric information and improve the security of user identity verification.

[0043] Figure 2 It is the second flowchart of a user verification method provided by an embodiment of the present invention. This embodiment is a refinement based on the above embodiment. The specific method can be as Figure 2 shown, and the method may include the following steps:

[0044] Step 201, obtain the handwritten image data of the user to be verified.

[0045] The handwritten image data includes the user identification to be verified, the handwriting platform information, the handwriting process parameters, and the handwritten image.

[0046] Step 202, verify the content of the handwritten image based on the user identification to be verified to obtain the content verification result of the handwritten image.

[0047] The content verification result is passed or not passed.

[0048] Step 203: Determine the feature indicators of the handwritten image based on the handwritten platform information, and extract features from the handwritten image according to the feature indicators and the handwritten process parameters to obtain the candidate features of the handwritten image.

[0049] Among them, the handwritten platform information is the information of the device and its related characteristics used by the user to be verified during handwritten. The handwritten platform information includes the type of handwritten device, such as a graphics tablet, the handwritten screen of a smart device, a computer mouse, or a paper document. It also includes the characteristics of the handwritten device, such as the sensor accuracy, pressure sensing range, sampling rate, and resolution of the handwritten device. It also includes the functions of the handwritten device, such as whether it supports pressure sensing, whether it supports multi-touch, and whether it supports trajectory recording. The feature indicators are used to guide the target handwriting verification model to extract features from the handwritten image. The handwritten process parameters are the dynamic data generated by the user to be verified during the writing process, and the handwritten process parameters can reflect the writing habits and styles of the user to be verified. The handwritten process parameters include force, stroke order, writing speed, writing trajectory, and dwell time, etc.

[0050] Specifically, different handwritten devices can correspond to different feature indicators. In this solution, optionally, determining the feature indicators of the handwritten image based on the handwritten platform information includes: when the handwritten platform is the first handwritten device, determining that the feature indicators include a pressure indicator, a speed indicator, and a stroke order indicator; when the handwritten platform is the second handwritten device, determining that the feature indicators include a speed indicator, a trajectory indicator, and a time indicator; when the handwritten platform is a paper document, determining that the feature indicators include a shape indicator and a contour indicator.

[0051] Among them, the first handwritten device includes a touch screen that can be handwritten and a graphics tablet, etc., and the second handwritten device includes a computer mouse, etc. The first handwritten device generally supports pressure sensing and stroke order recording. Therefore, when the handwritten platform is the first handwritten device, determining that the feature indicators include a pressure indicator, a speed indicator, and a stroke order indicator. When the handwritten platform is a mouse, the server can only record the mouse movement trajectory and dwell time. Therefore, when the handwritten platform is the second handwritten device, determining that the feature indicators include a speed indicator, a trajectory indicator, and a time indicator. When the handwritten platform is a paper document, the server can obtain a handwriting image including the handwritten content. Therefore, when the handwritten platform is a paper document, determining that the feature indicators include a shape indicator and a contour indicator.

[0052] Specifically, after determining the feature metrics, feature extraction is performed on the handwritten image according to the feature metrics and the handwritten process parameters, and the extracted features are determined as the candidate features of the handwritten image. Exemplarily, when the handwritten platform is the first handwritten device, the server extracts the pressure feature through the pressure of the stroke recorded by the sensor of the digitizing tablet, and calculates the writing speed through the timestamp and position coordinates uploaded by the digitizing tablet to obtain the speed feature. When the handwritten platform is the second handwritten device, the server calculates the writing speed through the trajectory and timestamp of the mouse movement to obtain the speed feature. By recording the trajectory of the mouse movement (the backward trajectory and the stop points), the trajectory feature is obtained. When the handwritten platform is a paper document, image processing algorithms such as the Canny operator, the Sobel operator, and the local binary pattern algorithm are used to extract the shape features of the handwritten content. The edge detection algorithm is used to extract the contour features of the handwritten content. By setting different feature metrics for different handwritten devices, the features of the handwriting can be captured and verified more accurately, ensuring that the verification process can adapt to different writing scenarios, thereby improving the accuracy of verification.

[0053] Step 204: Adjust the weights of the candidate features through the federated learning unit of the target handwriting verification model to obtain the weighted features of the handwritten image.

[0054] Among them, federated learning is a distributed machine learning technology. In this solution, the federated learning unit is used to perform weighted fusion and dynamic weight adjustment on the candidate features to adapt to the differences in handwriting features of different devices. In this solution, optionally, adjusting the weights of the candidate features through the federated learning unit of the target handwriting verification model to obtain the weighted features of the handwritten image includes: performing weighted fusion on the candidate features through the federated learning unit to obtain the fusion features corresponding to the candidate features; compressing the feature map of the fusion features to obtain the global feature vector; and performing dynamic weight adjustment on the global feature vector based on the fully connected layer of the federated learning unit to obtain the weighted features of the handwritten image.

[0055] Specifically, after obtaining the candidate features, the federated learning unit can dynamically calculate the weights of each feature through global pooling and fully connected layers, thereby enhancing the model's ability to perceive the importance of different feature channels. Using an attention mechanism module, such as through the SEblock (Squeeze-and-Excitation Block) technology, the candidate features are weighted and fused to obtain fused features. The SEblock can dynamically calculate the weights of each candidate feature through global pooling and fully connected layers, and perform this fusion on the candidate features according to their weights to obtain fused features. After obtaining the fused features, global pooling operations are used to compress the fused features to obtain a fixed-length global feature vector. Using a fully connected layer, according to the device type and user writing habits, the global feature vector is dynamically weighted and adjusted to obtain the weighted features of the handwritten image. Through the federated learning unit, the server can dynamically adjust the feature weights according to the handwritten device type, ensuring that the model can adapt to the handwriting features of different devices. Through federated learning, the target handwriting verification model can learn richer knowledge from the data of multiple devices, thereby improving the accuracy and robustness of the model.

[0056] Step 205: Obtain verification data based on the comparison unit and weighted features of the target handwriting verification model.

[0057] Among them, the comparison unit is used to calculate the similarity between the handwriting features in the handwriting image and the historical handwriting features of the user to be verified. In this solution, optionally, obtaining verification data based on the comparison unit and weighted features of the target handwriting verification model includes: calculating the similarity between the weighted features and the pre-determined historical handwriting features through the comparison unit; obtaining verification data based on the similarity and a pre-set similarity threshold.

[0058] Specifically, the target handwriting verification model is the handwriting verification model corresponding to the user to be verified, and the historical handwriting features are the historical handwriting features of the user to be verified. After obtaining the weighted features, the cosine similarity between the weighted features and the historical handwriting features can be calculated through the comparison unit. The higher the cosine similarity, the more similar the weighted features and the historical handwriting features are. Alternatively, the Manhattan distance or Hamming distance between the weighted features and the historical handwriting features can also be calculated, and the similarity between the weighted features and the historical handwriting features can be determined according to the Manhattan distance or Hamming distance. After obtaining the similarity between the weighted features and the historical handwriting features, the comparison unit can compare its similarity with the similarity threshold. If the similarity is higher than the similarity threshold, the verification result is determined to be verified passed. If the similarity is not higher than the similarity threshold, the verification result is determined to be verified not passed. After obtaining the verification result, verification data is generated based on the similarity and the verification result, etc. Through the comparison unit of the target handwriting verification model, the accuracy of handwriting verification is ensured, providing a reliable means of identity verification for application scenarios such as financial transactions.

[0059] Step 206: Determine the handwriting verification result based on the verification data, and display the handwriting verification result to the user to be verified.

[0060] In the technical solution of this embodiment, the handwritten image data of the user to be verified is obtained; the handwritten image data includes the user identification to be verified, handwritten platform information, handwritten process parameters, and handwritten images. Content verification is performed on the handwritten image based on the user identification to be verified to obtain the content verification result of the handwritten image; the content verification result is either passed or not passed. The characteristic index of the handwritten image is determined based on the handwritten platform information, and feature extraction is performed on the handwritten image according to the characteristic index and the handwritten process parameters to obtain the candidate features of the handwritten image. The weight of the candidate features is adjusted by the federated learning unit of the target handwriting verification model to obtain the weighted features of the handwritten image. Verification data is obtained based on the comparison unit and the weighted features of the target handwriting verification model. The handwriting verification result is determined based on the verification data, and the handwriting verification result is displayed to the user to be verified. In the technical solution of this embodiment, the characteristic index is determined according to different handwritten platform information, and static and dynamic features are combined to more comprehensively capture the user's writing habits and handwriting features. Even if an attacker can imitate the shape of the handwriting, it is difficult to simultaneously simulate dynamic features such as pressure and speed, improving the security and accuracy of user authentication. The weight of the candidate features is adjusted by the federated learning unit to dynamically adapt to the differences in handwriting features of different devices, ensuring that the model can authenticate the user more accurately according to the user's habits. That is, the technical solution of this embodiment can perform accurate uniqueness verification on the user, while ensuring that the user's biometric information is not leaked, improving the accuracy and security of user verification.

[0061] Figure 3 It is a flowchart of a method for determining a target handwriting verification model provided by an embodiment of the present invention. The specific method can be as Figure 3 shown, and the method may include the following steps:

[0062] Step 301: When the pre-established user verification model does not meet the preset conditions, obtain the training data of the user to be verified.

[0063] Among them, the training data includes the training image set of the user to be verified. The user verification model is the handwriting verification model that has not been trained corresponding to the user to be verified. The preset conditions are predetermined according to big data in the field and other factors, and are the performance indicators that the user verification model needs to meet in actual applications, such as verification accuracy, response time, or whether the loss function has converged, etc. The training image set is used to train the user verification model, and the training image set includes handwriting images written by the user on different devices. Specifically, when the user verification model does not meet the preset conditions, it means that the user verification model needs to be trained and optimized. Therefore, when the user verification model does not meet the preset conditions, collect the handwriting images written by the user to be verified on different handwriting devices, and annotate the collected handwriting images to ensure that the images correspond correctly to the user identity. Obtain the training image set of the user according to the annotated handwriting images.

[0064] Step 302: Optimize the parameters of the user verification model according to the training data of the user to be verified to obtain the target handwriting verification model.

[0065] Parameter optimization refers to adjusting the parameters of the user verification model to make the user verification model achieve better performance on the training data. The target handwriting verification model is the user verification model after parameter optimization, and the target handwriting verification model can more accurately identify the user's handwriting. After obtaining the training data, the user verification model can extract the features of the handwriting image, such as pressure, speed, and stroke order, etc., and obtain the verification data determined according to the features of the handwriting image output by the user verification model. Calculate the loss between the verification data and the real data. According to the backpropagation algorithm, adjust the model parameters of the user verification model according to this loss. Until the user verification model meets the preset conditions, the target handwriting verification model is obtained.

[0066] The technical solution of this embodiment, when the pre-established user verification model does not meet the preset conditions, obtains the training data of the user to be verified; the training data includes the training image set of the user to be verified, and the training image set includes the handwriting images of the user to be verified on various handwriting devices; optimizes the parameters of the user verification model according to the training data of the user to be verified to obtain the target handwriting verification model. The technical solution of this embodiment can obtain a target handwriting verification model that can accurately verify the user identity by optimizing the parameters of the user verification model, and improves the efficiency of user identity verification.

[0067] Figure 4 It is a schematic structural diagram of a user verification device provided by an embodiment of the present invention. This device is applicable to execute the user verification method provided by the embodiment of the present invention. As Figure 4 shown, this device may specifically include:

[0068] A data acquisition module 401, configured to acquire handwritten image data of a user to be verified; the handwritten image data includes an identifier of the user to be verified, handwritten platform information, handwritten process parameters, and a handwritten image.

[0069] A first verification module 402, configured to perform content verification on the handwritten image based on the identifier of the user to be verified, and obtain a content verification result of the handwritten image; the content verification result is either passed or not passed.

[0070] A second verification module 403, configured to, when the content verification result is passed, determine a target handwriting verification model based on the identifier of the user to be verified, and input the handwritten image data into the target handwriting verification model to obtain verification data output by the target handwriting verification model.

[0071] A result display module 404, configured to determine a handwriting verification result based on the verification data, and display the handwriting verification result to the user to be verified.

[0072] Optionally, the first verification module 402 is specifically configured to: determine target handwritten content corresponding to the user to be verified in a pre-determined user database based on the identifier of the user to be verified;

[0073] extract image content of the handwritten image through a pre-set image content extraction algorithm;

[0074] perform similarity matching between the image content and the target handwritten content to obtain a content verification result of the handwritten image.

[0075] Optionally, the second verification module 403 is specifically configured to: determine a feature index of the handwritten image based on the handwritten platform information, and perform feature extraction on the handwritten image according to the feature index and the handwritten process parameters to obtain candidate features of the handwritten image;

[0076] adjust weights of the candidate features through a federated learning unit of the target handwriting verification model to obtain weighted features of the handwritten image;

[0077] obtain the verification data based on a comparison unit of the target handwriting verification model and the weighted features.

[0078] Optionally, the second verification module 403 is further configured to: when the handwritten platform is a first handwritten device, determine that the feature index includes a pressure index, a speed index, and a stroke order index;

[0079] when the handwritten platform is a second handwritten device, determine that the feature index includes a speed index, a trajectory index, and a time index.

[0080] When the handwriting platform is a paper document, determining the feature indicators includes a shape indicator and a contour indicator.

[0081] Optionally, the second verification module 403 is further configured to: perform weighted fusion on the candidate features through the federated learning unit to obtain a fused feature corresponding to the candidate features;

[0082] Perform feature map compression on the fused feature to obtain a global feature vector;

[0083] Perform dynamic weight adjustment on the global feature vector based on the fully connected layer of the federated learning unit to obtain the weight feature of the handwritten image.

[0084] Optionally, the second verification module 403 is further configured to: calculate the similarity between the weight feature and a pre-determined historical handwriting feature through the comparison unit;

[0085] Obtain the verification data based on the similarity and a pre-set similarity threshold.

[0086] Optionally, the second verification module 403 is further configured to: when a pre-established user verification model does not meet a preset condition, obtain training data of the user to be verified; the training data includes a training image set of the user to be verified, and the training image set includes handwriting images of each handwriting device of the user to be verified;

[0087] Optimize the parameters of the user verification model according to the training data of the user to be verified to obtain the target handwriting verification model.

[0088] The user verification device provided in the embodiments of the present invention can execute the user verification method provided in any embodiment of the present invention, and has corresponding functional modules and beneficial effects for executing the method. The content not described in detail in this embodiment can be referred to the description in any method embodiment of the present invention.

[0089] The embodiments of the present invention further provide a computer program product.

[0090] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer program products, which can include one or more computer programs, the one or more computer programs being executable and / or interpretable on a programmable system including at least one programmable processor, the programmable processor being a special-purpose or general-purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0091] Figure 5 FIG. is a schematic structural diagram of an electronic device provided for an embodiment of the present invention. Refer to Figure 5 , Figure 5 The electronic device 12 shown is merely an example and should not impose any limitation on the functions and usage scope of the embodiments of the present application. As Figure 5 shown, the electronic device 12 is presented in the form of a general-purpose computing device. The components of the electronic device 12 can include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 connecting different system components (including the system memory 28 and the processing unit 16).

[0092] The bus 18 represents one or more of several types of bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. By way of example, these architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0093] The electronic device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 12, including volatile and non-volatile media, removable and non-removable media.

[0094] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 can be used for reading and writing on non-removable, non-volatile magnetic media ( Figure 5 not shown, commonly referred to as a "hard disk drive"). Although Figure 5 not shown in the figure, a disk drive for reading and writing on a removable non-volatile disk (such as a "floppy disk"), and an optical disk drive for reading and writing on a removable non-volatile optical disk (such as a CD-ROM, DVD-ROM or other optical media) can be provided. In these cases, each drive can be connected to the bus 18 through one or more data media interfaces. The memory 28 may include at least one program product having a set (such as at least one) of program modules configured to perform the functions of the embodiments of the present application.

[0095] A program / utility 40 having a set (at least one) of program modules 46 can be stored, for example, in the memory 28. Such program modules 46 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. The implementation of a network environment may be included in each or some combination of these examples. The program modules 46 generally perform the functions and / or methods in the embodiments described in the present application.

[0096] The electronic device 12 can also communicate with one or more external devices 14 (such as a keyboard, a pointing device, a display 24, etc.), and can also communicate with one or more devices that enable a user to interact with the electronic device 12, and / or communicate with any device that enables the electronic device 12 to communicate with one or more other computing devices (such as a network card, a modem, etc.). Such communication can be carried out through the input / output (I / O) interface 22. Moreover, the electronic device 12 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through the network adapter 20. As shown in the figure, the network adapter 20 communicates with other modules of the electronic device 12 through the bus 18. It should be understood that although Figure 5 not shown in the figure, other hardware and / or software modules can be used in combination with the electronic device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0097] The processing unit 16 executes various functional applications and data processing by running a program stored in the system memory 28, for example, implementing a user authentication method provided by an embodiment of the present invention: obtaining handwritten image data of a user to be authenticated; the handwritten image data includes an identifier of the user to be authenticated, handwritten platform information, handwritten process parameters, and a handwritten image; performing content verification on the handwritten image based on the identifier of the user to be authenticated to obtain a content verification result of the handwritten image; the content verification result is either passed or not passed; when the content verification result is passed, determining a target handwriting verification model based on the identifier of the user to be authenticated, and inputting the handwritten image data into the target handwriting verification model to obtain verification data output by the target handwriting verification model; determining a handwriting verification result based on the verification data, and presenting the handwriting verification result to the user to be authenticated.

[0098] Embodiments of the present invention provide a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements a user authentication method provided by all embodiments of the present invention: obtaining handwritten image data of a user to be authenticated; the handwritten image data includes an identifier of the user to be authenticated, handwritten platform information, handwritten process parameters, and a handwritten image; performing content verification on the handwritten image based on the identifier of the user to be authenticated to obtain a content verification result of the handwritten image; the content verification result is either passed or not passed; when the content verification result is passed, determining a target handwriting verification model based on the identifier of the user to be authenticated, and inputting the handwritten image data into the target handwriting verification model to obtain verification data output by the target handwriting verification model; determining a handwriting verification result based on the verification data, and presenting the handwriting verification result to the user to be authenticated. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, electronic devices, devices, or components of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the computer-readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or combined with an instruction-executing electronic device, device, or component.

[0099] A computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. The computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution electronic device, apparatus, or device.

[0100] The program code contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to wireless, wire, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0101] The computer program code for performing the operations of the present invention can be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network - including a local area network (LAN) or a wide area network (WAN) - or, alternatively, can be connected to an external computer (e.g., by using an Internet service provider to connect through the Internet).

[0102] Note that the above is only the preferred embodiment of the present invention and the technical principles applied. Those skilled in the art will understand that the present invention is not limited to the specific embodiments here. Various obvious changes, re-adjustments, and substitutions can be made by those skilled in the art without departing from the protection scope of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments. Without departing from the concept of the present invention, it can also include more other equivalent embodiments, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A user authentication method, characterized in that, The method includes: Obtaining the handwritten image data of the user to be verified; the handwritten image data includes the user identification to be verified, the handwritten platform information, the handwritten process parameters, and the handwritten image; Performing content verification on the handwritten image based on the user identification to be verified to obtain the content verification result of the handwritten image; the content verification result is passed or not passed; When the content verification result is passed, determining the target handwriting verification model based on the user identification to be verified, and inputting the handwritten image data into the target handwriting verification model to obtain the verification data output by the target handwriting verification model; Determining the handwriting verification result based on the verification data and presenting the handwriting verification result to the user to be verified.

2. The method according to claim 1, characterized in that, Performing content verification on the handwritten image based on the user identification to be verified to obtain the content verification result of the handwritten image, including: Determining the target handwritten content corresponding to the user to be verified in the pre-determined user database based on the user identification to be verified; Extracting the image content of the handwritten image through a pre-set image content extraction algorithm; Performing similarity matching between the image content and the target handwritten content to obtain the content verification result of the handwritten image.

3. The method according to claim 1, wherein Inputting the handwritten image data into the target handwriting verification model to obtain the verification data output by the target handwriting verification model, including: Determining the feature indicators of the handwritten image based on the handwritten platform information, and performing feature extraction on the handwritten image according to the feature indicators and the handwritten process parameters to obtain the candidate features of the handwritten image; Adjusting the weights of the candidate features through the federated learning unit of the target handwriting verification model to obtain the weighted features of the handwritten image; Obtaining the verification data based on the comparison unit of the target handwriting verification model and the weighted features.

4. The method according to claim 3, wherein Determining the feature indicators of the handwritten image based on the handwritten platform information, including: When the handwritten platform is the first handwritten device, determining that the feature indicators include a pressure indicator, a speed indicator, and a stroke order indicator; When the handwritten platform is the second handwritten device, determining that the feature indicators include a speed indicator, a trajectory indicator, and a time indicator; When the handwritten platform is a paper document, determining that the feature indicators include a shape indicator and a contour indicator.

5. The method according to claim 3, wherein Adjusting the weights of the candidate features through the federated learning unit of the target handwriting verification model to obtain the weighted features of the handwritten image, including: Performing weighted fusion on the candidate features through the federated learning unit to obtain the fusion features corresponding to the candidate features; Compressing the feature map of the fusion features to obtain a global feature vector; Performing dynamic weight adjustment on the global feature vector based on the fully connected layer of the federated learning unit to obtain the weighted features of the handwritten image.

6. The method according to claim 3, characterized in that Obtaining the verification data based on the comparison unit of the target handwriting verification model and the weighted features, including: Calculating the similarity between the weighted features and the pre-determined historical handwriting features through the comparison unit; Obtaining the verification data based on the similarity and the pre-set similarity threshold.

7. The method according to claim 1, characterized in that, Before obtaining the handwritten image data of the user to be verified, the method further includes: When the pre-established user verification model does not meet the preset conditions, obtain the training data of the user to be verified; the training data includes the training image set of the user to be verified, and the training image set includes the handwriting images of each handwriting device of the user to be verified. Optimize the parameters of the user verification model according to the training data of the user to be verified to obtain the target handwriting verification model.

8. A computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements a user verification method according to any one of claims 1-7.

9. An electronic device, the electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the user verification method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, it implements the user verification method according to any one of claims 1 to 7.

Images