Identity authentication method

Abstract

The invention belongs to the technical field of authentication, and provides an identity authentication method, which comprises the steps of personal account registration, risk standard setting, secondary verification trigger condition and verification pass condition setting, verification pair collection, detection data transmission, verification result judgment, risk control data collection, risk control recognition result judgment and risk standard updating. According to the invention, by opening the setting function of the secondary verification triggering condition and the verification passing condition, higher personalization is provided for user verification; a novel verification method is provided by collecting user audios and verification photos and setting verification pairs, and verification safety and specificity are improved; through risk control detection and risk standard updating, the defect of poor user experience caused by frequent triggering of a verification mechanism is reduced, and the universality of the authentication method is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of authentication, and in particular to an identity authentication method. Background Art

[0002] Authentication, or identity verification, is to verify whether the user has the authority to access or operate relevant data from the server. When sending a request, the corresponding verification parameters must usually be included to ensure that the request has access rights and returns the required data. For example, you must enter a username and password to log in to a website, and some services require logging in before they can be performed. In order to reduce the problem of malicious attacks on servers and leakage of personal sensitive information, major platforms have set up a two-step verification mechanism. Common two-step verification technologies include: SMS verification code, email verification code, identity verification application, hardware security key, biometric technology, push notification, security questions, and device identification. Among them, in order to prevent robot programs such as crawlers from accessing the server, the platform usually uses text verification codes, picture verification codes, audio verification codes, sliding verification codes, behavior analysis, time analysis, invisible verification codes, question-and-answer verification codes, and interactive verification codes.

[0003] However, with the development of artificial intelligence, the above technologies are becoming increasingly easy to hack, resulting in reduced security. In addition, the existing verification methods are basically automatically generated by the system, resulting in less personalization for users. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide an identity authentication method, which provides higher personalization for user verification by opening the setting function of secondary verification trigger conditions and verification pass conditions; by collecting user audio and verification photos and setting verification pairs, a new verification method is provided to improve the security and specificity of verification; through risk control detection and risk standard updates, the defects of frequent triggering of the verification mechanism leading to poor user experience are reduced, thereby improving the versatility of the authentication method.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] An identity authentication method, comprising:

[0007] Register a platform account and get a personal account;

[0008] Setting risk criteria;

[0009] Based on the personal account, set secondary verification trigger conditions and verification pass conditions;

[0010] Entering several sets of one-to-one corresponding user audios and verification photos, and binding the corresponding user audios and verification photos to obtain several sets of verification pairs;

[0011] When the secondary verification trigger condition is triggered, a number of the user audio segments, a preset number of the verification photos, and a number of randomly generated interference pictures are randomly sent to the client to be detected;

[0012] Collect the matching results of the client to be detected, analyze the matching results and the verification pass condition, and obtain the verification result;

[0013] When the verification result is successful, allowing the client to be detected to access the server;

[0014] After the client to be detected logs into the server, risk control detection is performed on the behavior of the client to be detected to obtain risk control data;

[0015] When the risk control data exceeds the risk standard, the client to be detected is repeatedly verified by forcibly triggering the secondary verification trigger condition to obtain a risk control identification result;

[0016] When the risk control identification result is that there is no risk issue, the risk standard is increased.

[0017] Preferably, the calculation formula of the risk standard is:

[0018]

[0019] Wherein, R is the risk standard; R0 is the basic risk value; and n is the number of repeated verifications.

[0020] Preferably, the secondary verification triggering conditions include: logging in from an uncommon address, changing the login device, frequent login attempts, accessing private files, and frequent file deletions.

[0021] Preferably, the verification pass conditions include: matching accuracy, number of matching errors and matching time.

[0022] Preferably, several groups of one-to-one corresponding user audios and verification photos are recorded, and the corresponding user audios and verification photos are bound to obtain several groups of verification pairs, including:

[0023] Collect the target user’s uploaded audio and photos;

[0024] Noise reduction and compression are performed on the uploaded audio to obtain the user audio;

[0025] Performing noise reduction and resolution standardization on the uploaded photo to obtain the verification photo;

[0026] According to the matching scheme of the target user, the user audio and the verification photo corresponding to each other are packaged and stored to obtain the verification pair.

[0027] Preferably, when the secondary verification trigger condition is triggered, a number of the user audio segments, a preset number of the verification photos and a number of randomly generated interference pictures are randomly sent to the client to be detected, including:

[0028] When the secondary verification triggering condition is triggered, randomly selecting a number of groups of verification pairs;

[0029] Randomly selecting a number of the user audio segments and a preset number of the verification photos from the selected verification pairs;

[0030] Randomly generate a number of interference pictures according to the difference between the selected user audio and the verification photo;

[0031] The selected user audio, the verification photo and the interference picture are sent to the client to be detected.

[0032] Preferably, collecting the matching results of the client to be detected, analyzing the matching results and the verification pass conditions to obtain the verification results, includes:

[0033] Receiving feedback information sent by the client to be detected;

[0034] Parsing the feedback information to obtain the matching result;

[0035] Using the verification pass condition to judge the matching result, to obtain a judgment result;

[0036] When all the judgment results are in compliance, the verification result is determined to be successful;

[0037] When any of the judgment results is not in compliance, the verification result is determined to be a verification failure.

[0038] Preferably, after the client to be detected logs into the server, risk control detection is performed on the behavior of the client to be detected to obtain risk control data, including:

[0039] After the client to be detected logs into the server, the behavior information of the client to be detected is recorded;

[0040] According to a preset risk behavior score set, the total value of the behavior information is calculated to obtain the risk control data.

[0041] The present invention discloses the following technical effects:

[0042] The present invention provides an identity authentication method, which solves the defect that the verification methods of the existing methods are all fixedly set by opening the setting function of the secondary verification trigger condition and the verification pass condition, and realizes the personalized setting of users according to their own needs; by collecting user audio and verification photos and setting verification pairs, it solves the defect that the traditional verification method is easily cracked by artificial intelligence, and realizes a more secure verification function; through risk control detection and risk standard update, it solves the defect that the verification mechanism of the existing method is frequently triggered, and realizes the update of risk standards according to the increase in the number of successful verifications. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0044] Figure 1 A schematic diagram of an identity authentication process provided by an embodiment of the present invention;

[0045] Figure 2 An identity authentication flow chart provided for an embodiment of the present invention;

[0046] Figure 3 A schematic diagram of the verification pair acquisition process provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] The purpose of the present invention is to provide an identity authentication method, which provides higher personalization for user authentication by opening the setting function of secondary verification trigger conditions and verification pass conditions; provides a new verification method by collecting user audio and verification photos and setting verification pairs to improve the security and specificity of verification; reduces the defect of poor user experience caused by frequent triggering of the verification mechanism through risk control detection and risk standard updates, and improves the versatility of the authentication method.

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0050] Figure 1A schematic diagram of an identity authentication process provided by an embodiment of the present invention, Figure 2 The identity authentication flow chart provided by the embodiment of the present invention is as follows: Figure 1 and Figure 2 As shown, the present invention provides an identity authentication method, comprising:

[0051] Step 100: Register a platform account and obtain a personal account;

[0052] Step 200: Setting risk criteria;

[0053] Step 300: Setting secondary verification trigger conditions and verification pass conditions based on the personal account;

[0054] Step 400: inputting several sets of one-to-one corresponding user audios and verification photos, and binding the corresponding user audios and verification photos to obtain several sets of verification pairs;

[0055] Step 500: When the secondary verification trigger condition is triggered, a number of user audio segments, a preset number of verification photos, and a number of randomly generated interference pictures are randomly sent to the client to be detected;

[0056] Step 600: Collect the matching results of the client to be detected, analyze the matching results and the verification pass condition, and obtain the verification result;

[0057] Step 700: When the verification result is successful, allow the client to be detected to access the server;

[0058] Step 800: After the client to be detected logs into the server, risk control detection is performed on the behavior of the client to be detected to obtain risk control data;

[0059] Step 900: When the risk control data exceeds the risk standard, the client to be detected is repeatedly verified by forcibly triggering the secondary verification trigger condition to obtain a risk control identification result;

[0060] Step 1000: When the risk control identification result is no risk issue, increase the risk standard.

[0061] Specifically, the calculation formula of the risk standard is:

[0062]

[0063] Wherein, R is the risk standard; R0 is the basic risk value; and n is the number of repeated verifications.

[0064] Optionally, the secondary verification triggering conditions include: logging in from an uncommon address, changing the login device, frequent login attempts, accessing private files, and frequent file deletion.

[0065] Preferably, the verification pass conditions include: matching accuracy, number of matching errors and matching time.

[0066] refer to Figure 3 , input several sets of one-to-one corresponding user audio and verification photos, and bind the corresponding user audio and verification photos to obtain several sets of verification pairs, including:

[0067] Step 401: Collect the target user's uploaded audio and uploaded photos;

[0068] Step 402: performing noise reduction and compression on the uploaded audio to obtain the user audio;

[0069] Step 403: performing noise reduction and resolution standardization on the uploaded photo to obtain the verification photo;

[0070] Step 404: According to the matching scheme of the target user, the corresponding user audio and the verification photo are packaged and stored to obtain the verification pair.

[0071] Preferably, in order to ensure the normal operation and speed of the system authentication function, the bit rate of the uploaded audio shall not be lower than 192kbps, and the resolution of the uploaded photos shall not be lower than 720p; when the bit rate of the uploaded audio exceeds 192kbps, the system will convert it into 192kbps format; when the uploaded photos exceed 1080p, the system will convert them into 1080p format.

[0072] Furthermore, common denoising techniques include: DnCNN, FFDNet, CBDNet, RIDNet, PMRID, SID. The main features of the DnCNN network are: the network is divided into three parts, the first part is Conv+Relu (one layer), the second part is Conv+BN+Relu (several layers), and the third part is Conv (one layer), and the number of network layers is 17 or 20. The network learns the image residual, that is, the difference between the noisy image and the noise-free image, and the loss function uses MSE. The main features of the FFDNet network are: the structure of the network is consistent with that of DnCNN, but the input and output are different. The network input is four sub-images obtained by downsampling the original noisy image and a noise level image generated by the parameter σ input by the user. The output is four denoised sub-images, and the final denoised image is obtained by upsampling. The loss function used is still MSE; because the input of the network contains a user-controlled parameter, the algorithm is better than DnCNN in adapting to different noises. The main features of the CBDNet network are: the structure of the network is divided into two parts: the first part is a five-layer fully convolutional network for noise estimation and noise level map, and the second part is UnNet with residuals, which is different from FFDNet, for noise reduction; an asymmetric loss function is designed, the main purpose is to eliminate asymmetric sensitivity. The so-called asymmetric sensitivity refers to noise reduction algorithms such as BM3D and FFDNet. When the input noise reduction parameter σ is small, the noise reduction effect is poor. When the input noise reduction parameter σ is large, although the texture is lost, the noise reduction effect is still relatively good. Based on the above description, this embodiment selects CBDNet technology as a method for denoising the uploaded photos.

[0073] Preferably, the method for denoising the uploaded audio in this embodiment is DPCRN_DNS3. DPCRN_DNS3 uses the TensorFlow framework and combines libraries such as numpy, matplotlib, librosa and soxfile to build a powerful and flexible model. It mainly relies on DPCRN, an innovative neural network architecture that combines the advantages of convolutional and recurrent neural networks to effectively suppress environmental noise while maintaining the naturalness and integrity of speech. Its advantages are: the DPCRN model achieves a good balance between noise reduction effect and computational efficiency; in addition to offline testing, the project also includes real-time inference scripts, which can achieve real-time speech enhancement on the device; it supports quantization and pruning of TensorFlow Lite, and the model can be used on resource-constrained platforms such as mobile devices; it provides a concise command line interface to facilitate model training, testing and real-time processing; it is open source and has clear reference guidance.

[0074] Specifically, when the secondary verification trigger condition is triggered, a number of user audio segments, a preset number of verification photos, and a number of randomly generated interference pictures are randomly sent to the client to be detected, including:

[0075] When the secondary verification triggering condition is triggered, randomly selecting a number of groups of verification pairs;

[0076] Randomly selecting a number of the user audio segments and a preset number of the verification photos from the selected verification pairs;

[0077] Randomly generate a number of interference pictures according to the difference between the selected user audio and the verification photo;

[0078] The selected user audio, the verification photo and the interference picture are sent to the client to be detected.

[0079] Further, the matching results of the client to be detected are collected, and the matching results and the verification pass conditions are analyzed to obtain the verification results, including:

[0080] Receiving feedback information sent by the client to be detected;

[0081] Parsing the feedback information to obtain the matching result;

[0082] Using the verification pass condition to judge the matching result, to obtain a judgment result;

[0083] When all the judgment results are in compliance, the verification result is determined to be successful;

[0084] When any of the judgment results is not in compliance, the verification result is determined to be a verification failure.

[0085] Preferably, after the client to be detected logs into the server, risk control detection is performed on the behavior of the client to be detected to obtain risk control data, including:

[0086] After the client to be detected logs into the server, the behavior information of the client to be detected is recorded;

[0087] According to a preset risk behavior score set, the total value of the behavior information is calculated to obtain the risk control data; the risk behavior score set can be set according to user needs.

[0088] Preferably, this embodiment provides a similar picture screening method based on KNN for providing the interference picture to the client to be detected. The design idea is as follows:

[0089] Step S1: Collect the verification photos and user audio of all users on the entire platform;

[0090] Step S2: filtering the verification photo and the user audio; this step is used to filter the verification photo with poor display quality and the user audio with too much interference frequency;

[0091] Step S3: combining the filtered verification photo and the user audio according to the matching information of the verification pair, and discarding redundant unpaired verification photo and user audio;

[0092] Step S4: extracting text information of the user audio, extracting keywords based on the text information, and combining the keywords with the corresponding verification photos to obtain training samples;

[0093] Step S5: randomly divide the training samples into a training set and a test set (7:3);

[0094] Step S6: construct a k-neighbor network;

[0095] Step S7: constructing a loss function and an optimization function;

[0096] Step S8: training the k-nearest neighbor network to obtain a classification model.

[0097] Specifically, kNN (k-nearest neighbor algorithm) is a basic and simple classification algorithm. As a supervised learning, the KNN model needs labeled training data. The category of a new sample is determined by the k training sample points closest to the new sample according to the classification decision rule. The k-nearest neighbor method (kNN) is a basic classification and regression method; a model based on labeled training data; and a supervised learning algorithm. The three key points of the basic approach are: first, determining the distance metric; second, the selection of the k value (finding the k instance points closest to the estimated point in the training set); and third, the classification decision rule. In the classification task, the "voting method" can be used, that is, selecting the most frequently appearing labeled category in these k instances as the prediction result; in the regression task, the "average method" can be used, that is, the average value of the real-valued output labels of these k instances is used as the prediction result; and weighted averaging or weighted voting can also be performed based on the distance, and the closer the distance, the greater the weight of the instance. The k-nearest neighbor method does not have an explicit learning process. It is a well-known representative of lazy learning. This type of learning technology simply saves samples during the training phase, with zero training time overhead, and processes them after receiving test samples.

[0098] Optionally, the main method of k-nearest neighbor clustering analysis is: initially randomly given K (hyperparameter) cluster centers, these K cluster centers are randomly selected from the sample points. The sample points to be classified are divided into clusters according to the nearest neighbor principle. After the division is completed, the centroid of each cluster is recalculated by the averaging method to determine the new cluster center. Iterate until the moving distance of the cluster center is less than a given value. The evaluation criterion of the algorithm is: the sum of squared errors criterion. When the sum of squared errors reaches the optimal (small) value, the class can be as compact as possible and the clusters can be separated as much as possible. The loss function is a non-convex function, so there are many minimum values, which may not always fall to the global optimum, but may fall to the local optimum. The loss function is used for global optimization during iteration. The remaining data is clustered according to the silhouette coefficient; the calculation formula of the silhouette coefficient is:

[0099]

[0100] Among them, s is the silhouette coefficient; b is the sum of the distances between the data to be assigned and all the points in the next nearest cluster; a is the sum of the distances between the data to be assigned and other points in the same cluster; max() is the maximum value function. The range of the silhouette coefficient is (-1, 1). The closer the value is to 1, the more similar the sample is to the sample in its own cluster, and the less similar it is to the samples in other clusters. When the sample point is more similar to the sample points outside the cluster, the silhouette coefficient is negative. When the silhouette coefficient is 0, it means that the similarity between the two clusters is the same, and the two clusters should be one cluster.

[0101] Furthermore, when the authentication method provided in this embodiment is running, it is mandatory that the user audios sent to the client to be detected for identity authentication all come from the feature account to be logged in, and the number of user audios is greater than the verification photos. Ultimately, it is required that the number of pictures received by the client to be detected is twice the number of audios.

[0102] Specifically, when the authentication method provided in this embodiment is running, the number of user audio and verification photos sent to the client to be detected is set according to the verification pass condition set in advance by the user, the difference between the two is calculated, and the classification model is used to generate a corresponding number of interference pictures.

[0103] Furthermore, the tool used to extract the text information of the user audio is FFmpeg. FFmpeg is a set of open source computer programs that can be used to record, convert digital audio and video, and convert them into streams. It adopts LGPL or GPL license. It provides a complete solution for recording, converting and streaming audio and video. It contains a very advanced audio / video codec library libavcodec. In order to ensure high portability and codec quality, many codes in libavcodec are developed from scratch. FFmpeg is developed under the Linux platform, but it can also be compiled and run in other operating system environments, including Windows, Mac OS X, etc. This project was first initiated by Fabrice Bellard and was mainly maintained by Michael Niedermayer from 2004 to 2015. Many FFmpeg developers come from the MPlayer project, and FFmpeg is currently placed on the server of the MPlayer project team. The name of the project comes from the MPEG video coding standard, and the "FF" in front stands for "FastForward". The FFmpeg coding library can use GPU acceleration. FFmpeg is a set of open source computer programs that can be used to record, convert digital audio and video, and convert them into streams. It includes the leading audio / video encoding library libavcodec, etc. libavformat: used for the generation and parsing of various audio and video encapsulation formats, including obtaining the information required for decoding to generate the decoding context structure and read audio and video frames; libavcodec: used for various types of sound / image encoding and decoding; libavutil: contains some common tool functions; libswscale: used for video scene scaling and color mapping conversion; libpostproc: used for post-effect processing; ffmpeg: a tool provided by the project, which can be used for format conversion, decoding or TV card real-time encoding, etc.; ffsever: an HTTP multimedia real-time broadcast streaming server; ffplay: is a simple player that uses the ffmpeg library to parse and decode and display through SDL. This embodiment provides a basic pseudo code for video and audio file extraction based on ffmpeg:

[0104] ffmpeg-i source_video.avi-vn-ar44100-ac 2-ab 192-fmp3 sound.mp3;

[0105] Among them, source_video.avi is the original video file, and it can also be other types of video files; for example: mp4, etc.; 192-f is the audio bit rate, that is, the audio bit rate of the audio file is set to 192kb / s; mp3 is the target output format; sound.mp3 is the output audio file name.

[0106] The beneficial effects of the present invention are as follows:

[0107] The present invention provides higher personalization for user verification by opening the setting function of secondary verification trigger conditions and verification pass conditions; provides a new verification method by collecting user audio and verification photos and setting verification pairs, which improves the security and specificity of verification; reduces the defect of poor user experience caused by frequent triggering of verification mechanism through risk control detection and risk standard update, and improves the versatility of authentication method

[0108] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0109] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. An identity authentication method, characterized in that: include: Register a platform account and get a personal account; Setting risk criteria; Based on the personal account, set secondary verification trigger conditions and verification pass conditions; Entering several sets of one-to-one corresponding user audios and verification photos, and binding the corresponding user audios and verification photos to obtain several sets of verification pairs; When the secondary verification trigger condition is triggered, a number of the user audio segments, a preset number of the verification photos, and a number of randomly generated interference pictures are randomly sent to the client to be detected; Collect the matching results of the client to be detected, analyze the matching results and the verification pass condition, and obtain the verification result; When the verification result is successful, allowing the client to be detected to access the server; After the client to be detected logs into the server, risk control detection is performed on the behavior of the client to be detected to obtain risk control data; When the risk control data exceeds the risk standard, the client to be detected is repeatedly verified by forcibly triggering the secondary verification trigger condition to obtain a risk control identification result; When the risk control identification result is that there is no risk issue, the risk standard is increased.

2. An identity authentication method according to claim 1, characterized in that: The risk criterion is calculated as follows: Wherein, R is the risk standard; R0 is the basic risk value; and n is the number of repeated verifications.

3. An identity authentication method according to claim 1, characterized in that: The secondary verification triggering conditions include: logging in from an uncommon address, changing the login device, frequent login attempts, accessing private files, and frequent file deletions.

4. An identity authentication method according to claim 1, characterized in that: The verification pass conditions include: matching accuracy, number of matching errors and matching time.

5. An identity authentication method according to claim 1, characterized in that: Enter several sets of one-to-one corresponding user audio and verification photos, and bind the corresponding user audio and verification photos to obtain several sets of verification pairs, including: Collect the target user’s uploaded audio and photos; Noise reduction and compression are performed on the uploaded audio to obtain the user audio; Performing noise reduction and resolution standardization on the uploaded photo to obtain the verification photo; According to the matching scheme of the target user, the user audio and the verification photo corresponding to each other are packaged and stored to obtain the verification pair.

6. An identity authentication method according to claim 1, characterized in that: When the secondary verification trigger condition is triggered, a number of user audio segments, a preset number of verification photos, and a number of randomly generated interference pictures are randomly sent to the client to be detected, including: When the secondary verification triggering condition is triggered, randomly selecting a number of groups of verification pairs; Randomly selecting a number of the user audio segments and a preset number of the verification photos from the selected verification pairs; Randomly generate a number of interference pictures according to the difference between the selected user audio and the verification photo; The selected user audio, the verification photo and the interference picture are sent to the client to be detected.

7. An identity authentication method according to claim 1, characterized in that: Collecting the matching results of the client to be detected, analyzing the matching results and the verification pass conditions to obtain the verification results, including: Receiving feedback information sent by the client to be detected; Parsing the feedback information to obtain the matching result; Using the verification pass condition to judge the matching result, to obtain a judgment result; When all the judgment results are in compliance, the verification result is determined to be successful; When any of the judgment results is not in compliance, the verification result is determined to be a verification failure.

8. An identity authentication method according to claim 1, characterized in that: After the client to be detected logs into the server, the behavior of the client to be detected is subjected to risk control detection to obtain risk control data, including: After the client to be detected logs into the server, the behavior information of the client to be detected is recorded; According to a preset risk behavior score set, the total value of the behavior information is calculated to obtain the risk control data.

Images