Password verification method, device and equipment and computer readable storage medium

Abstract

The invention discloses a password verification method, device and equipment and a computer readable storage medium. The method is applied to the wearable intelligent device and comprises the steps that a password input interface is displayed in response to a trigger operation of password verification, the password input interface comprises a plurality of text-type characters and numerical-type characters corresponding to the text-type characters in a one-to-one mode, and the text-type characters have different pronunciations; in response to voice input of a plurality of target characters selected by the user from the plurality of text-type characters, determining voice information input by the user; performing voice recognition on the voice information, and determining a text type character password; converting the text type character password into a numeric type character password based on the corresponding relationship between the text type character and the numeric type character; and sending the numeric character password to a password verification system, so that the password verification system verifies the numeric character password. According to the embodiment of the invention, in the password verification process, the password security and the password input convenience can be ensured at the same time.

Description

Technical Field

[0001] The present application belongs to the field of intelligent visual payment technology, and in particular, relates to a password verification method, device, equipment and computer-readable storage medium. Background Art

[0002] In the smart visual payment scenario, wearable smart devices can interact with users through eye tracking, gestures, voice, etc., collect user payment intentions, payment amounts, and user identity information, and verify user identity information. If the verification is successful, the payment is completed based on the user's payment intentions and payment amount. Among them, user identity information verification can be achieved through traditional digital password verification. As a classic, secure and time-tested verification method, digital passwords have shown strong compatibility and stability in smart visual payment scenarios.

[0003] Traditional digital password verification methods usually verify the password entered by the user on a physical keyboard or a virtual keyboard. However, the traditional digital password verification method has the risk of password leakage during the password input process, resulting in low password security. In addition, due to the limited screen area of ​​the physical keyboard or virtual keyboard and the compact keyboard layout, it is easy to cause accidental touches, making it difficult for users to input.

[0004] Therefore, there is an urgent need for a password verification method that can simultaneously ensure password security and password input convenience. Summary of the invention

[0005] The embodiments of the present application provide a password verification method, apparatus, device, computer-readable storage medium, and computer program product, which can ensure password security and password input convenience during the password verification process.

[0006] In a first aspect, an embodiment of the present application provides a password verification method, which is applied to a wearable smart device, and the method includes:

[0007] In response to a triggering operation of password verification, a password input interface is displayed, the password input interface includes a virtual keyboard, the virtual keyboard includes a plurality of text characters and their one-to-one corresponding numeric characters, the correspondence between the text characters and the numeric characters is generated based on system time, a target user identifier and a random number, the target user identifier is a user identifier having a binding relationship with the wearable smart device, and the plurality of text characters have different pronunciations;

[0008] In response to a user's voice input of selecting a plurality of target characters from the plurality of text-type characters, determining voice information input by the user;

[0009] Performing voice recognition on the voice information to determine a text-based character password;

[0010] Based on the correspondence between the text characters and the numeric characters, converting the text character password into a numeric character password;

[0011] The numerical character password is sent to a password verification system so that the password verification system verifies the numerical character password.

[0012] In a possible implementation, after responding to the triggering operation of password verification and before displaying the password input interface, the method further includes:

[0013] Acquire a text character set, a plurality of numeric characters, the system time, the target user identifier, and the random number, wherein the text character set includes a plurality of the text characters;

[0014] Encrypting the plurality of numeric characters based on the system time, the target user identifier and the random number to obtain a plurality of first encrypted information;

[0015] Acquire the plurality of text characters with different pronunciations respectively corresponding to the plurality of first encrypted information from the text character set;

[0016] Mapping the plurality of numeric characters to the plurality of text characters to obtain a correspondence between the plurality of numeric characters and the plurality of text characters;

[0017] The virtual keyboard is determined based on a correspondence between the plurality of numeric characters and the plurality of text characters.

[0018] In a possible implementation, the encrypting the multiple numeric characters based on the system time, the target user identifier, and the random number to obtain multiple first encrypted information includes:

[0019] Generate a target key based on the system time, the target user identifier and the random number through a hash algorithm;

[0020] Based on the Advanced Encryption Standard algorithm, the target key is used to encrypt the multiple numeric characters to obtain the multiple first encrypted information.

[0021] In a possible implementation, the converting the text-type character password into a numeric-type character password based on the correspondence between the text-type character and the numeric-type character includes:

[0022] Based on the correspondence between the text characters and the first encryption information, converting the text character password into an encryption information sequence, wherein the encryption information sequence includes the plurality of first encryption information;

[0023] Based on the Advanced Encryption Standard algorithm, the encrypted information sequence is decrypted to obtain the numerical character password.

[0024] In a possible implementation, sending the numeric character password to a password verification system includes:

[0025] Encrypting the numerical character password to obtain second encrypted information;

[0026] The second encrypted information is sent to the password verification system.

[0027] In a possible implementation, in response to a user's voice input of selecting a plurality of target characters from the plurality of text characters, determining the voice information input by the user includes:

[0028] authenticating the user and obtaining a verification result;

[0029] When the verification result is that the verification is passed, receiving a voice input of the user selecting a plurality of target characters from the plurality of text characters;

[0030] In response to the voice input, voice information of the user input is determined.

[0031] In a possible implementation, performing voice recognition on the voice information to determine the text-based character password includes:

[0032] Inputting the speech information into a speech recognition model;

[0033] The voice information is subjected to voice recognition through the voice recognition model to determine the text-based character password.

[0034] In a possible implementation manner, the speech information is a time domain signal, and the step of inputting the speech information into a speech recognition model includes:

[0035] Converting the speech information from the time domain signal to a frequency domain signal by fast Fourier transform;

[0036] Extracting spectral features from the frequency domain signal;

[0037] The spectral features are input into the speech recognition model.

[0038] In a possible implementation, the speech recognition model includes a convolutional layer, an attention layer, a pooling layer, and a fully connected layer; and performing speech recognition on the speech information through the speech recognition model to determine the text-based character password includes:

[0039] Performing a convolution operation on the frequency spectrum feature through the convolution layer to obtain a first feature map;

[0040] Downsampling the first feature map through the pooling layer to obtain a second feature map;

[0041] Determining weights of features in the second feature map by the attention layer;

[0042] Mapping the features in the second feature map to a plurality of text characters and probabilities corresponding to the plurality of text characters respectively based on weights of the features in the second feature map through an activation function in the fully connected layer;

[0043] The text character password is determined based on the plurality of text characters and the probabilities respectively corresponding to the plurality of text characters.

[0044] In a possible implementation, before inputting the speech information into the speech recognition model, the method further includes:

[0045] Acquire multiple training samples, each of which includes a first voice information sample and a first character label corresponding to it;

[0046] Inputting the first voice information sample into an initial voice recognition model to perform voice recognition on the first voice information sample through the initial voice recognition model to obtain a first predicted character;

[0047] Determining a loss function value according to the similarity between the first predicted character and the first character label;

[0048] When the loss function value does not meet the preset training stop condition, the first model parameter of the initial speech recognition model is adjusted based on the back propagation algorithm, and the process of inputting the first speech information sample into the initial speech recognition model is returned to execute, so as to perform speech recognition on the first speech information sample through the initial speech recognition model to obtain a first predicted character, until the loss function value meets the preset training stop condition, thereby obtaining the trained speech recognition model.

[0049] In a possible implementation, after obtaining the trained speech recognition model, the method further includes:

[0050] Acquire multiple test samples, each of which includes a second voice information sample and a corresponding second character label;

[0051] Inputting the second voice information sample into the voice recognition model to perform voice recognition on the second voice information sample through the voice recognition model to obtain a second predicted character;

[0052] Determining a character error rate based on similarities between a plurality of the second predicted characters and their corresponding first character labels;

[0053] When the character error rate is greater than a preset threshold, the second model parameters of the speech recognition model are adjusted, and the step of inputting the second speech information sample into the speech recognition model is returned to execute so as to perform speech recognition on the second speech information sample through the speech recognition model to obtain a second predicted character, until the character error rate is less than or equal to the preset threshold.

[0054] In a second aspect, an embodiment of the present application provides a password verification device, which is applied to a wearable smart device, and the device includes:

[0055] a display module, configured to display a password input interface in response to a trigger operation of password verification, wherein the password input interface includes a virtual keyboard, wherein the virtual keyboard includes a plurality of text characters and corresponding numeric characters thereof, wherein the correspondence between the text characters and the numeric characters is generated based on system time, a target user identifier and a random number, wherein the target user identifier is a user identifier having a binding relationship with the wearable smart device, and wherein the plurality of text characters have different pronunciations;

[0056] A first determining module, configured to determine the voice information input by the user in response to the user's voice input of selecting a plurality of target characters from the plurality of text-type characters;

[0057] A second determination module is used to perform voice recognition on the voice information to determine the text-based character password;

[0058] A conversion module, used for converting the text-type character password into a numeric character password based on the correspondence between the text-type character and the numeric character;

[0059] The sending module is used to send the numerical character password to the password verification system so that the password verification system verifies the numerical character password.

[0060] In a third aspect, an embodiment of the present application provides an electronic device, the device comprising: a processor and a memory storing computer program instructions;

[0061] When the processor executes the computer program instructions, the processor implements any possible implementation method of the first aspect described above.

[0062] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, implements a method in any possible implementation method of the first aspect described above.

[0063] In a fifth aspect, an embodiment of the present application provides a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes a method as any possible implementation method in the first aspect above.

[0064] The embodiment of the present application can ensure the security of the correspondence between text characters and numerical characters by displaying multiple text characters and their one-to-one corresponding numerical characters in the password input interface of the wearable smart device, and can also enable the user who inputs the password to clearly understand the encryption information corresponding to the actual numerical character password (i.e., the text character password). Since multiple text characters have different pronunciations, the voice information input by the user is determined in response to the voice input of the user selecting multiple target characters from multiple text characters, the voice information is voice recognized, the text character password is determined, and based on the correspondence between the text characters and the numerical characters, the text character password is converted into a numerical character password, so that the password can be input through voice interaction. On the one hand, the user can enter the password without manually operating the password keyboard, which improves the convenience of password input; on the other hand, since the user who does not wear the wearable smart device cannot know the correspondence between the text characters and the numerical characters, the password security can be guaranteed. In addition, by generating a correspondence between text characters and numeric characters based on system time, target user identification and random numbers, that is, each time in response to a trigger operation for password verification, the displayed virtual keyboard is different, which increases the difficulty of guessing or cracking the password, and further improves the password security. In this way, by receiving the encrypted password information (i.e., the text character password) input by the user's voice, and sending the decrypted numeric character password to the password verification system, so that the password verification system verifies the numeric character password, the password security and the convenience of password input can be guaranteed during the password verification process. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0066] Figure 1This is a schematic diagram of an embodiment of the present application providing a method of scanning a merchant's QR code based on smart glasses and implementing a payment process;

[0067] Figure 2 It is a flowchart of a password verification method provided in an embodiment of the present application;

[0068] Figure 3 is a schematic diagram of a password verification method provided in an embodiment of the present application;

[0069] Figure 4 It is a structural schematic diagram of a password verification device provided in an embodiment of the present application;

[0070] Figure 5 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0071] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0072] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0073] In addition, the acquisition, storage, use, and processing of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0074] As described in the background technology section, in the intelligent visual payment scenario, wearable smart devices can interact with users through eye tracking, gestures, voice, etc., collect user payment intentions, payment amounts and user identity information, and verify user identity information. If the verification is passed, the payment is completed based on the user's payment intentions and payment amount.

[0075] The schematic diagram of a payment process based on scanning a merchant QR code with smart glasses provided in the embodiment of the present application can be as follows: Figure 1 shown.

[0076] like Figure 1 As shown, smart glasses can collect user payment intention and payment amount by scanning the merchant's QR code. Afterwards, in response to the user's confirmation operation of the user's payment intention and payment amount, the smart glasses can collect the user's identity information and send the user's payment intention, payment amount and user identity information to the background server. The background server can parse the user's payment intention through the vertical big model in the financial payment field, determine the key payment information, and verify the user's identity information. If the identity authentication is passed, the key payment information and payment amount are encapsulated into a prescribed protocol format and sent to the payment server to implement the payment process through the payment server, and the payment result is returned to the smart glasses, the acquiring institution and the merchant.

[0077] In the prior art, user identity information verification can be achieved through traditional digital password verification methods, as well as biometric recognition methods such as fingerprint recognition, face recognition, voiceprint recognition, and iris recognition.

[0078] Specifically, existing user identity information verification methods include:

[0079] (1) Enter a digital password through a traditional keyboard and verify the digital password. The traditional keyboard includes a physical keyboard on a standard computer and a virtual keyboard on a mobile device.

[0080] (2) Biometric identification method: If supported by specific banks or specific terminal manufacturers, users can authenticate their identities through biometric information such as fingerprints and faces, thereby eliminating the need to enter a payment password.

[0081] The prior art has the following shortcomings or problems:

[0082] (1) Traditional keyboard input: There is a risk of password leakage. Passwords entered directly through a traditional keyboard are easily recorded by hackers, and peepers may infer the password by observing the keyboard tapping position. In addition, using a traditional keyboard to enter passwords on a mobile device is prone to accidental touches due to the limited screen area and compact keyboard layout. Some wearable smart devices cannot be installed with a traditional keyboard, making it difficult for users to enter passwords.

[0083] (2) Biometric identification method:

[0084] Fingerprint recognition: Fingerprint recognition technology is widely used due to its convenience and high recognition rate, but its popularity and application are limited by the support of the issuing banks. Currently, not all banks have access to or fully support fingerprint verification functions. This lack of uniformity in verification methods may lead to compatibility issues when users use wearable smart devices for payment, thus affecting the universality and user experience of fingerprint recognition. In addition, the security of fingerprint recognition also faces challenges, such as the risk of fingerprint templates being stolen or copied, and the recognition failure that may occur in certain special circumstances (such as injured or wet fingers), which limits its wider application.

[0085] Face recognition: In the field of digital payment and security verification, face recognition technology is relatively mature and widely used, but it still has limitations as a single-factor verification condition. Because face recognition technology is easily affected by factors such as light, angle, and obstructions, it is difficult to ensure sufficient security by relying solely on face recognition as a means of verification. Therefore, in digital scenarios, face recognition often needs to be used in combination with other verification methods (such as passwords, SMS verification codes, etc.). In addition, when face recognition technology is applied to visual intelligent payment scenarios, especially when the camera is integrated into the glasses worn by the user, it may not only be difficult to collect due to the wearer's behavior (such as turning the head, lowering the head), but may also cause users to worry about privacy leaks, because continuous facial capture may infringe on the user's personal privacy.

[0086] Voiceprint recognition: As an identity authentication technology based on sound features, voiceprint recognition is unique and non-replicable, but its technology is still immature and has not yet reached the stability and accuracy requirements of production-level applications. Voiceprint recognition is easily affected by factors such as environmental noise, recording equipment quality, and speaker emotional changes, resulting in a high recognition error rate. In addition, the collection, storage, and processing of voiceprint data also face strict privacy protection and legal compliance requirements, all of which restrict the widespread application and commercialization of voiceprint recognition technology.

[0087] Iris recognition: Iris recognition technology is considered one of the cutting-edge technologies in the field of biometrics due to its extremely high accuracy and stability. However, its widespread use is also accompanied by potential hazards to human health. Long-term and frequent use of iris recognition equipment may cause discomfort or damage to the human eye, such as eye fatigue and dryness caused by strong light stimulation. In addition, the high cost of iris recognition technology, including investment in equipment purchase, maintenance, and user education, also limits its large-scale promotion in the general consumer market. Therefore, although iris recognition technology has unique advantages, it is still necessary to weigh the relationship between its security, convenience and health risks in actual applications.

[0088] In summary, in the scenario of intelligent visual payment, the traditional digital password input method is facing the dual challenges of convenience and security. Although biometric recognition methods such as fingerprint recognition, face recognition, and voiceprint recognition have made it possible to improve the payment experience, the application of these methods has certain limitations. Specifically, although fingerprint recognition is convenient, its popularization requires the support of the issuing bank. Not all banks across the country are compatible with this function, and it lacks universal applicability. Face recognition cannot be used as a means of verification alone in digital payment, and when the camera is integrated into wearable devices such as glasses, the user's facial features become difficult to capture. Voiceprint recognition technology is still in its early stages of development and has not yet reached the maturity of large-scale commercial use. As for iris recognition, despite its high accuracy, its potential impact on human health makes the risks of long-term use cannot be ignored.

[0089] Therefore, as a classic, secure and time-tested verification method, digital passwords have shown strong compatibility and stability in the field of intelligent visual payment, especially in cross-border payment scenarios. It can not only effectively ensure transaction security, but also achieve seamless connection in different payment environments and systems, becoming an indispensable security verification method in the current technological context.

[0090] However, the traditional digital password verification method has the risk of password leakage during the password input process, resulting in low password security. In addition, due to the limited screen area of ​​the physical keyboard or virtual keyboard and the compact keyboard layout, it is easy to cause accidental touches and difficult for users to input.

[0091] Therefore, there is an urgent need for a password verification method that can simultaneously ensure password security and password input convenience.

[0092] Thus, in order to solve the problems of the prior art, the embodiments of the present application provide a password verification method, apparatus, device, computer-readable storage medium and computer program product. The password verification method can be applied to the scenario of using a wearable smart device for payment.

[0093] The following first introduces the password verification method provided in the embodiment of the present application.

[0094] Figure 2 FIG. 1 is a flow chart of a password verification method provided by an embodiment of the present application. The password verification method can be executed by a wearable smart device. Figure 2 As shown, the password verification method provided in the embodiment of the present application includes the following steps:

[0095] S210, in response to a triggering operation of password verification, displaying a password input interface, the password input interface including a virtual keyboard, the virtual keyboard including a plurality of text characters and their one-to-one corresponding numeric characters, the correspondence between the text characters and the numeric characters being generated based on system time, a target user identifier and a random number, the target user identifier being a user identifier having a binding relationship with the wearable smart device, and the plurality of text characters having different pronunciations;

[0096] S220, in response to a user's voice input of selecting a plurality of target characters from a plurality of text-type characters, determining voice information input by the user;

[0097] S230, performing voice recognition on the voice information to determine the text character password;

[0098] S240, based on the correspondence between text characters and numeric characters, convert the text character password into a numeric character password;

[0099] S250: Send the numeric character password to the password verification system, so that the password verification system verifies the numeric character password.

[0100] The embodiment of the present application can ensure the security of the correspondence between text characters and numerical characters by displaying multiple text characters and their one-to-one corresponding numerical characters in the password input interface of the wearable smart device, and can also enable the user who inputs the password to clearly understand the encryption information corresponding to the actual numerical character password (i.e., the text character password). Since multiple text characters have different pronunciations, the voice information input by the user is determined in response to the voice input of the user selecting multiple target characters from multiple text characters, the voice information is voice recognized, the text character password is determined, and based on the correspondence between the text characters and the numerical characters, the text character password is converted into a numerical character password, so that the password can be input through voice interaction. On the one hand, the user can enter the password without manually operating the password keyboard, which improves the convenience of password input; on the other hand, since the user who does not wear the wearable smart device cannot know the correspondence between the text characters and the numerical characters, the password security can be guaranteed. In addition, by generating a correspondence between text characters and numeric characters based on system time, target user identification and random numbers, that is, each time in response to a trigger operation for password verification, the displayed virtual keyboard is different, which increases the difficulty of guessing or cracking the password, and further improves the password security. In this way, by receiving the encrypted password information (i.e., the text character password) input by the user's voice, and sending the decrypted numeric character password to the password verification system, so that the password verification system verifies the numeric character password, the password security and the convenience of password input can be guaranteed during the password verification process.

[0101] The specific implementation methods of the above steps are introduced below.

[0102] In some embodiments, in S210, in a payment scenario, the triggering operation of password verification may be an operation for the user to confirm payment. In addition, the password input interface may be an interface presented in a display panel of a wearable smart device using projection technology. The password input interface may include a virtual keyboard. The virtual keyboard may include a plurality of text characters and their one-to-one corresponding numerical characters, and the plurality of text characters have different pronunciation codes. Specifically, the virtual keyboard may include a plurality of virtual controls, each of which may correspond to a text character and a numerical character corresponding to the text character. Among them, the text character may be at least one of Chinese characters and English letters. The numerical characters may include the numbers 0-9.

[0103] In addition, the correspondence between text characters and numeric characters in the virtual keyboard can be generated based on system time, target user identification and random numbers. The system time can be the system time of the wearable smart device. The target user identification can be a user identification that has a binding relationship with the wearable smart device.

[0104] In the embodiment of the present application, since the system time is constantly changing and the password generated each time may be different, the correspondence between text characters and numeric characters is generated based on the system time, the target user identifier and the random number, that is, each time in response to the trigger operation of password verification, the displayed virtual keyboard is different, which increases the difficulty of guessing or cracking the password and improves the password security.

[0105] Based on this, in order to increase the difficulty of guessing or cracking the password and improve the security of the password, in some embodiments, after the above-mentioned triggering operation in response to the password verification and before displaying the password input interface, the method may further include:

[0106] Obtain a text character set, multiple numeric characters, system time, target user ID, and random number, where the text character set includes multiple text characters;

[0107] Based on the system time, the target user identifier and the random number, multiple numeric characters are encrypted to obtain multiple first encrypted information;

[0108] Acquire a plurality of text characters with different pronunciations corresponding to the plurality of first encrypted information respectively from the text character set;

[0109] Mapping multiple numeric characters to multiple text characters to obtain a correspondence between the multiple numeric characters and the multiple text characters;

[0110] A virtual keyboard is determined based on a correspondence between a plurality of numeric characters and a plurality of text characters.

[0111] Here, the text character set may include multiple Chinese characters and English letters. In combination with scenario requirements, the types and quantities of text characters in the text character set may be continuously adjusted, and symbols may be added to the text character set to cope with more complex security requirements and further improve password security. In addition, the multiple numeric characters may be numbers in a numeric keypad. The multiple numeric characters may include numbers 0-9. In addition, the random number may be generated by a random number generator.

[0112] After obtaining multiple numeric characters, system time, target user identifier and random number, multiple numeric characters can be encrypted based on system time, target user identifier and random number to obtain multiple first encrypted information. Among them, each numeric character can correspond to a first encrypted information. After obtaining the first encrypted information, the first encrypted information can be used as an index or seed to randomly select multiple text characters corresponding to the multiple first encrypted information from the text character set. The multiple text characters obtained at this time may have the same pronunciation, or the pronunciations may be different. If any two text characters have the same pronunciation, multiple text characters corresponding to the multiple first encrypted information can be re-selected until the pronunciations of the multiple text characters are different. Among them, the way to determine whether the pronunciations of multiple text characters are the same can be to pre-assign a pronunciation code corresponding to the pronunciation to each text character, and determine whether the pronunciations are the same by checking the pronunciation code.

[0113] In addition, in order to ensure that the multiple text characters corresponding to the multiple first encrypted information have different pronunciations, multiple sub-text character sets can be generated in advance based on the text character set, and the sub-text character sets can include multiple text characters with different pronunciations, and the number of text character sets is the same as the number of first confidential information. In this way, obtaining multiple text characters with different pronunciations corresponding to the multiple first encrypted information in the text character set can specifically include: selecting a sub-text character set from the multiple sub-text character sets, and randomly determining the correspondence between the multiple text characters in the sub-text character set and the multiple first encrypted information.

[0114] After determining multiple text characters with different pronunciations, multiple numeric characters in the numeric keyboard can be mapped to the above-mentioned multiple text characters to obtain a correspondence between the multiple numeric characters and the multiple text characters, and a virtual keyboard is determined based on the correspondence between the multiple numeric characters and the multiple text characters, and the virtual keyboard is displayed on the display panel of the wearable smart device.

[0115] The embodiment of the present application first encrypts multiple numeric characters based on system time, target user identifier and random number to obtain multiple first encrypted information, then uses the first encrypted information as index or seed, randomly selects text characters corresponding to multiple numeric characters in the text character set, and determines the virtual keyboard based on the correspondence between multiple numeric characters and multiple text characters. That is, through the dynamic encryption mapping mechanism, it can be achieved that each time the trigger operation of password verification is responded, the displayed virtual keyboard is different, which increases the difficulty of guessing or cracking the password and improves the security of the password. In addition, by first generating the first encrypted information of the numeric character, and then using the first encrypted information as index or seed to determine the text character corresponding to the numeric character, compared with directly determining the correspondence between the numeric character and the text character, the difficulty of password reverse deduction is increased, and the security of the password is improved.

[0116] Based on this, in order to ensure the randomness and security of the correspondence between numeric characters and text characters, in some embodiments, the above-mentioned encryption of multiple numeric characters based on system time, target user identifier and random number to obtain multiple first encrypted information may specifically include:

[0117] Generate a target key based on system time, target user ID and random number through a hash algorithm;

[0118] Based on the Advanced Encryption Standard algorithm, a plurality of numeric characters are encrypted using a target key to obtain a plurality of first encrypted information.

[0119] Here, the hash algorithm may be, for example, a secure hash algorithm (SHA). By using a hash algorithm (such as SHA-256), taking the system time, the target user ID, and a random number generator as input, a sufficiently long target key may be generated for subsequent cryptographic mapping generation, ensuring that each operation is unique.

[0120] In addition, the wearable smart device may include a secure storage environment. The secure storage environment may include a hardware security module (HSM) and a secure memory area. Among them, the hardware security module is a computer hardware device used to protect and manage the keys used by the strong authentication system and provide related cryptographic operations. After the target key is generated, the target key can be stored in the secure storage environment to ensure the security of the target key. In addition, by adding a tampering mechanism to the wearable smart device, it can be ensured that the firmware and operating system of the wearable smart device are not tampered with.

[0121] Exemplarily, by encrypting multiple numeric characters (such as numbers 0-9) using a target key based on the Advanced Encryption Standard (AES) algorithm, first encrypted information corresponding to the numbers 0-9 can be obtained.

[0122] On this basis, when the first encrypted information is used as an index or seed and multiple text characters corresponding to multiple first encrypted information are randomly selected from the text character set, if any two text characters have the same pronunciation, the above-mentioned process of generating the target key and after generating the target key can be returned until the pronunciations of the multiple text characters are all different.

[0123] The embodiment of the present application uses a hash algorithm and an Advanced Encryption Standard algorithm to encrypt multiple numeric characters based on system time, target user identification and random numbers, thereby ensuring the randomness and security of the correspondence between subsequently generated numeric characters and text characters.

[0124] In some embodiments, in S220, after the password input interface is displayed in the wearable smart device, the user can see multiple text characters and their one-to-one corresponding numeric characters, and input the text character password by voice. Among them, the text character password may include multiple target characters and the order between the multiple target characters. Since the user who does not wear the wearable smart device has no way of knowing the correspondence between the text characters and the numeric characters, the method of the user inputting the text character password by voice will not disclose the password.

[0125] As an example, a wearable smart device may be equipped with a microphone, which may collect voice information input by a user.

[0126] Since other people's voice information may also be collected during the process of collecting voice information, in order to improve the success rate of password verification, in some embodiments, the above S220 may specifically include:

[0127] Authenticate the user and obtain the verification result;

[0128] When the verification result is that the verification is passed, receiving a voice input of the user selecting a plurality of target characters from a plurality of text characters;

[0129] In response to the voice input, voice information input by the user is determined.

[0130] Here, the identity authentication may include at least one of fingerprint recognition verification and personal identification number (PIN code) two-factor verification.

[0131] The embodiment of the present application receives a text-based character password input by the user's voice when the user identity verification is passed, thereby ensuring that the password obtained in the subsequent password decryption, password transmission, etc. is the password input by the user wearing the wearable smart device, thereby improving the success rate of password verification.

[0132] In some embodiments, in S230, after the voice information is collected, voice recognition may be performed on the voice information to determine the text character password input by the user.

[0133] In order to ensure the accuracy of the text-based character password, in some embodiments, the above S230 may specifically include:

[0134] Input the speech information into the speech recognition model;

[0135] The voice information is recognized through the voice recognition model to determine the text-based character password.

[0136] Here, the speech recognition model can be a pre-trained neural network model for speech recognition. The speech recognition model can specifically be a convolutional neural network (CNN) model based on an attention mechanism. The speech recognition model may include a convolutional layer, an attention layer, a pooling layer, and a fully connected layer. Among them, the attention layer can follow the convolutional layer, the pooling layer can follow the attention layer, and the fully connected layer can follow the pooling layer. In addition, the relative position relationship between the convolutional layer, the attention layer, the pooling layer, and the fully connected layer can also be: the pooling layer follows the convolutional layer, the attention layer follows the pooling layer, and the fully connected layer can follow the attention layer, which is not limited here.

[0137] As an example, the above-trained speech recognition model and password security protection control can be deployed in a wearable smart device. After the user inputs speech information based on the virtual keyboard, the speech information can be collected in real time through the microphone, and the speech information can be input into the speech recognition model for speech recognition. After the speech recognition obtains the text-based character password, the text-based character password can be fed back to the password security protection control in the form of text or other forms.

[0138] The embodiment of the present application performs speech recognition on speech information through the above-mentioned speech recognition model to determine the text-based character password, thereby ensuring the accuracy of the text-based character password.

[0139] In order to further improve the accuracy of text-based character passwords, before inputting the voice information into the speech recognition model, the background noise in the voice information can be removed by using a filter technology, and then the continuous voice signal can be divided into multiple shorter frames (such as 20-30 milliseconds) for subsequent processing. In addition, each frame can be weighted using a window function such as a Hamming window or a Hanning window to reduce the discontinuity of the frame boundary.

[0140] On this basis, since the collected speech signal is a time domain signal, and the phonemes, pitch, timbre and other characteristics of the speech signal can be more easily analyzed and identified in the frequency domain, each of the above windowed frames can be subjected to a Fast Fourier Transform (FFT) to decompose the time domain signal into the sum of a series of sinusoidal waves of different frequencies to obtain a frequency domain signal, and then perform speech recognition based on the frequency domain signal.

[0141] Based on this, in order to further improve the accuracy of the text-based character password, in some embodiments, the above-mentioned inputting of voice information into the voice recognition model may specifically include:

[0142] Through fast Fourier transform, the speech information is converted from time domain signal to frequency domain signal;

[0143] Extract spectral features from frequency domain signals;

[0144] The spectral features are fed into the speech recognition model.

[0145] Here, through the fast Fourier transform, in addition to converting the voice information from the time domain signal to the frequency domain signal, the noise reduction processing of the voice signal can also be realized at the same time. Since the noise is usually distributed in a wide frequency range, and the voice signal is concentrated in a specific frequency band, the fast Fourier transform can also suppress the noise by filtering out the noise frequency components in the frequency domain. In addition, after obtaining the frequency domain signal, the clarity and recognizability of the voice signal can be further enhanced by adjusting the amplitude and phase of each frequency component in the frequency domain.

[0146] In this way, by extracting spectral features from frequency domain signals and inputting the spectral features into a speech recognition model, and then performing speech recognition on the spectral features through the speech recognition model to determine the text-based character password, the accuracy of the text-based character password can be further improved.

[0147] On this basis, in order to further improve the accuracy of the text-based character password, in some embodiments, the above-mentioned voice recognition of the voice information through the voice recognition model to determine the text-based character password may specifically include:

[0148] Performing a convolution operation on the spectral features through a convolution layer to obtain a first feature map;

[0149] Downsample the first feature map through a pooling layer to obtain a second feature map;

[0150] Determine the weights of the features in the second feature map through the attention layer;

[0151] Through the activation function in the fully connected layer, based on the weights of the features in the second feature map, the features in the second feature map are mapped to a plurality of text characters and probabilities corresponding to the plurality of text characters respectively;

[0152] A text character password is determined based on a plurality of text characters and the probabilities that the plurality of text characters respectively correspond to.

[0153] Here, the convolution layer can use multiple convolution kernels (also called filters) to perform convolution operations on the spectral features to generate a first feature map. The first feature map may include local feature information of the spectral features. The pooling layer can be used to downsample the first feature map to reduce the dimension and amount of calculation of the data while retaining key information to obtain a second feature map. The attention layer can dynamically adjust the weights of different time steps or feature channels so that the model pays more attention to important speech segments, and finally obtains the weights of the features in the second feature map. In addition, the fully connected layer can be used to map the features in the second feature map to a text-type character password. Specifically, the activation function can be a softmax function, and the fully connected layer can convert the outputs of the convolution layer and the pooling layer into a probability distribution through the softmax function to represent the possibility of different categories (i.e., which text-type character), and determine the text-type character password based on the probability distribution.

[0154] The embodiment of the present application can further improve the accuracy of the text-based character password by determining the weights of the features in the second feature map based on the attention layer, and determining the text-based character password based on the weights of the features in the second feature map.

[0155] Based on this, in order to determine the text-based character password through the speech recognition model, in some embodiments, before inputting the speech information into the speech recognition model, the method may further include:

[0156] Acquire multiple training samples, each training sample includes a first voice information sample and a first character label corresponding thereto;

[0157] Inputting the first voice information sample into an initial voice recognition model to perform voice recognition on the first voice information sample through the initial voice recognition model to obtain a first predicted character;

[0158] Determining a loss function value according to a similarity between the first predicted character and the first character label;

[0159] When the loss function value does not meet the preset training stop condition, the first model parameter of the initial speech recognition model is adjusted based on the back propagation algorithm, and the execution returns to input the first speech information sample into the initial speech recognition model, so as to perform speech recognition on the first speech information sample through the initial speech recognition model to obtain a first predicted character, until the loss function value meets the preset training stop condition, and a trained speech recognition model is obtained.

[0160] Here, the first character label can be a text-type character label, and the first character label can include Chinese characters and English letters. The first voice information sample can be a voice information entered by the user based on the first character label. The initial speech recognition model can be a speech recognition model before training with multiple training samples. In addition, the risk of model overfitting can be reduced by adding an L2 policy to the loss function of the model. In addition, during the model training process, the generalization ability of the model can be improved by data enhancement, regularization technology (dropout), and the like.

[0161] The embodiment of the present application obtains a trained speech recognition model by training an initial speech recognition model based on a first speech information sample and its corresponding first character label, and can determine a text-based character password through the speech recognition model.

[0162] Based on this, in order to improve the accuracy and robustness of speech recognition, in some embodiments, after obtaining the trained speech recognition model as described above, the method may further include:

[0163] Acquire multiple test samples, each test sample includes a second voice information sample and its corresponding second character label;

[0164] Inputting the second voice information sample into the voice recognition model to perform voice recognition on the second voice information sample through the voice recognition model to obtain a second predicted character;

[0165] determining a character error rate based on similarities between the plurality of second predicted characters and their corresponding first character labels;

[0166] When the character error rate is greater than the preset threshold, the second model parameters of the speech recognition model are adjusted, and the execution returns to input the second speech information sample into the speech recognition model so as to perform speech recognition on the second speech information sample through the speech recognition model to obtain a second predicted character until the character error rate is less than or equal to the preset threshold.

[0167] Here, the second character tag may be a text-type character tag, and the second character tag may include Chinese characters and English letters. The second voice information sample may be voice information entered by the user based on the second character tag. In addition, the character error rate may include at least one of a word error rate (Word Error Rate, WER) and a character error rate (Character Error Rate, CER).

[0168] The embodiments of the present application evaluate the performance of the model by calculating indicators such as word error rate (WER) or character error rate (CER), and tune the model according to the evaluation results, so as to improve the accuracy and robustness of speech recognition.

[0169] In some embodiments, in S240, as described above, after determining the text character password, the text character password can be transmitted to the password security protection control. The password security protection control can perform an inverse operation based on the correspondence between the text character and the numeric character to map the text character password back to the original numeric character password.

[0170] In addition, as described above, the method for generating the correspondence between text characters and numeric characters can be: first, through the Advanced Encryption Standard algorithm, use the target key to encrypt multiple numeric characters to obtain multiple first encrypted information, and then use the first encrypted information as an index or seed to randomly obtain text characters corresponding to the multiple first encrypted information in the text character set to obtain the correspondence between text characters and numeric characters.

[0171] Therefore, in some embodiments, the above S240 may specifically include:

[0172] Based on the correspondence between the text characters and the first encrypted information, converting the text character password into an encrypted information sequence, where the encrypted information sequence includes a plurality of first encrypted information;

[0173] Based on the Advanced Encryption Standard algorithm, the encrypted information sequence is decrypted to obtain a numerical character password.

[0174] Here, the password security protection control may store an advanced encryption standard algorithm. The algorithm for decrypting the encrypted information sequence may be the same as the algorithm for encrypting the numeric characters. By first converting the text character password into an encrypted information sequence, and then decrypting the encrypted information sequence based on the advanced encryption standard algorithm, the text character password can be restored to a numeric character password.

[0175] In some embodiments, in S250, if the numeric password is a payment password, the password verification system may be a backend system of the issuing bank. After obtaining the numeric character password, the numeric character password may be sent to the password verification system so that the password verification system verifies and processes the numeric character password.

[0176] Based on this, in order to further improve the password security, in some embodiments, the above S250 may specifically include:

[0177] Encrypting the numeric character password to obtain second encrypted information;

[0178] The second encrypted information is sent to the password verification system.

[0179] Here, the second encrypted information can be specifically obtained by encrypting the numerical character password using the Transport Layer Security (TLS) protocol or the Secure Sockets Layer (SSL) protocol. In addition, after obtaining the second encrypted information, the second encrypted information can also be sent to the password verification system through a secure channel.

[0180] In this way, the password security can be further improved by re-encrypting the numerical password during the password transmission process.

[0181] In addition, the embodiments of the present application can also record key operations such as key access, password decryption and transmission requests during the password verification process for the purpose of auditing and monitoring.

[0182] In order to better describe the entire solution, some specific examples are given based on the above embodiments.

[0183] For example, Figure 3 As shown, the wearable smart device in the embodiment of the present application may include a dynamic encryption mapping module, a voice recognition module and a password security protection control. Among them, the dynamic encryption mapping module can respond to the triggering operation of password verification to call up the password input interface; generate a dynamic encryption mapping table (including the correspondence between text characters and the numerical characters) based on the system time, the target user identifier and the random number; based on the dynamic encryption mapping table, map multiple numerical characters in the numeric keyboard to multiple text characters to obtain a virtual keyboard; and display the virtual keyboard.

[0184] It should be noted that, based on the above method, a new and unique encrypted mapping table can be generated each time a user logs in or trades, and the input from the numeric keypad is converted into a randomly assigned Chinese character or letter sequence in real time and displayed on the screen of the wearable smart device. To ensure security and uniqueness, the mapping relationship between each number and the corresponding Chinese character / letter is different in each operation and has a different pronunciation.

[0185] In addition, the speech recognition module can receive the speech information input by the user; the CNN speech recognition algorithm based on the self-attention mechanism performs Fourier transform on the speech information to obtain the frequency domain signal; the spectrum features are extracted from the frequency domain signal; the speech recognition model performs speech recognition based on the spectrum features to obtain the text character password. Among them, the speech recognition model can automatically learn and extract the deep information in these spectrum features to achieve accurate recognition of the speech signal.

[0186] In addition, the password security protection control can receive a text character password; restore the text character password to a numeric character password; encrypt the numeric character password to obtain a second encrypted information; and send the second encrypted information to a password verification system to ensure the security and integrity of the password information.

[0187] In summary, this application proposes a random encryption mapping password input method based on voice interaction. In the payment scenario of wearable smart devices, this method uses projection technology to present the password input interface on the display panel of the wearable smart device, and adopts a dynamic random mapping mechanism to map the numeric keyboard to Chinese characters or English letters. It receives the encrypted characters (i.e., text-type character password) input by the user through efficient voice interaction, restores the text-type character password to the original numeric character password, and sends the numeric character password to the issuing bank for verification while ensuring full security encryption.

[0188] The invention points are as follows:

[0189] (1) Projection technology presents the password input interface on the display panel of the wearable smart device: Projection technology is used to enable the display panel of the wearable smart device to present the password input interface. This function developed for the wearable smart device uses its own display panel to carry the password input related display, which is different from the traditional operation on other fixed screens. It is an invention point that is closely integrated with the hardware characteristics of smart glasses.

[0190] (2) Password input on wearable smart devices based on voice interaction: The feature of wearable smart devices that facilitate voice interaction is fully utilized to allow users to input encrypted characters through voice, which changes the conventional method of manually inputting passwords. This allows users to complete password-related operations conveniently and efficiently on such wearable devices with relatively limited operation. This closely fits the usage of wearable smart devices and is a unique feature of the invention.

[0191] (3) Dynamic random mapping mechanism is applied to password input of wearable smart devices: The dynamic random mapping mechanism is applied to wearable smart devices, the numeric keyboard is mapped to Chinese characters or English letters, and the encrypted mapping conversion of password input is realized in the display and interaction scenarios of wearable smart devices. This special password conversion mechanism is innovated in combination with wearable smart devices, making password input more secure and innovative.

[0192] Therefore, this application has the following advantages:

[0193] (1) By randomly mapping the numeric keypad to Chinese characters or letters, the mapping relationship is different each time it is used, which increases the difficulty of guessing or cracking the password. The dynamic random mapping mechanism can flexibly adjust the mapping rules.

[0194] (2) Use efficient voice interaction to input passwords. Users do not need to manually operate the keyboard, especially on wearable smart devices (such as smart glasses). This method is more convenient and natural, reducing the complexity and inconvenience of manual operation. The built-in voice recognition algorithm analyzes the user's voice in combination with the context. The user can accurately recognize the user's voice input after the user inputs the encrypted Chinese characters or English letters through voice.

[0195] (3) The encrypted voice input is accepted through the password security protection control, and the same encryption algorithm is used to restore it to the original digital password. The encrypted state is always maintained during the transmission to the issuing bank for verification, ensuring the security of the password information during transmission.

[0196] Based on the password verification method applied to the wearable smart device provided in the above embodiment, accordingly, the present application also provides a specific implementation of the password verification device applied to the wearable smart device. Please refer to the following embodiments.

[0197] like Figure 4 As shown, the password verification device 400 provided in the embodiment of the present application includes the following modules:

[0198] The display module 410 is used to display a password input interface in response to a trigger operation of password verification, wherein the password input interface includes a virtual keyboard, wherein the virtual keyboard includes a plurality of text characters and their one-to-one corresponding numeric characters, wherein the correspondence between the text characters and the numeric characters is generated based on the system time, the target user identifier and the random number, wherein the target user identifier is a user identifier having a binding relationship with the wearable smart device, and the plurality of text characters have different pronunciations;

[0199] A first determination module 420, configured to determine the voice information input by the user in response to the voice input of the user selecting a plurality of target characters from a plurality of text-type characters;

[0200] The second determination module 430 is used to perform speech recognition on the speech information to determine the text-based character password;

[0201] A conversion module 440, configured to convert a text-type character password into a numeric-type character password based on a correspondence between text-type characters and numeric-type characters;

[0202] The sending module 450 is used to send the numeric character password to the password verification system so that the password verification system verifies the numeric character password.

[0203] The password verification device 400 is described in detail below, as follows:

[0204] In some embodiments, the password verification device 400 may further include:

[0205] A first acquisition module is used to acquire a text character set, a plurality of numeric characters, a system time, a target user identifier, and a random number after responding to a trigger operation of password verification and before displaying a password input interface, wherein the text character set includes a plurality of text characters;

[0206] An encryption module, used to encrypt multiple numeric characters based on system time, target user identifier and random number to obtain multiple first encrypted information;

[0207] A second acquisition module is used to acquire, from a text character set, a plurality of text characters with different pronunciations corresponding to the plurality of first encrypted information;

[0208] A mapping module, used for mapping a plurality of numeric characters to a plurality of text characters, and obtaining a correspondence between the plurality of numeric characters and the plurality of text characters;

[0209] The third determining module is used to determine the virtual keyboard based on the correspondence between the multiple numeric characters and the multiple text characters.

[0210] In some embodiments, the encryption module may specifically include:

[0211] A generation submodule, for generating a target key based on system time, target user ID and random number by using a hash algorithm;

[0212] The first encryption submodule is used to encrypt multiple numeric characters using a target key based on an Advanced Encryption Standard algorithm to obtain multiple first encrypted information.

[0213] In some embodiments, the conversion module 440 may specifically include:

[0214] A conversion submodule, used for converting the text character password into an encrypted information sequence based on the correspondence between the text character and the first encrypted information, wherein the encrypted information sequence includes a plurality of first encrypted information;

[0215] The decryption submodule is used to decrypt the encrypted information sequence based on the Advanced Encryption Standard algorithm to obtain a numerical character password.

[0216] In some embodiments, the sending module 450 may specifically include:

[0217] The second encryption submodule is used to encrypt the numerical character password to obtain second encrypted information;

[0218] The sending submodule is used to send the second encrypted information to the password verification system.

[0219] In some embodiments, the first determining module 420 may specifically include:

[0220] The verification submodule is used to authenticate the user and obtain the verification result;

[0221] A receiving submodule, for receiving a voice input of a user selecting a plurality of target characters from a plurality of text characters when the verification result is that the verification is passed;

[0222] The first determination submodule is used to determine the voice information input by the user in response to the voice input.

[0223] In some embodiments, the second determining module 430 may specifically include:

[0224] A first input submodule, used for inputting speech information into a speech recognition model;

[0225] The second determination submodule is used to perform speech recognition on the speech information through a speech recognition model to determine the text-based character password.

[0226] In some embodiments, the voice information is a time domain signal. Based on this, the first input submodule may specifically include:

[0227] A conversion unit, used for converting the speech information from a time domain signal into a frequency domain signal by fast Fourier transform;

[0228] An extraction unit, used for extracting spectrum features from frequency domain signals;

[0229] The input unit is used to input the spectral features into the speech recognition model.

[0230] In some embodiments, the speech recognition model includes a convolutional layer, an attention layer, a pooling layer, and a fully connected layer. Based on this, the second determination submodule may specifically include:

[0231] A convolution unit, configured to perform a convolution operation on the spectrum feature through a convolution layer to obtain a first feature map;

[0232] A sampling unit, used for downsampling the first feature map through a pooling layer to obtain a second feature map;

[0233] A first determining unit, configured to determine weights of features in a second feature map through an attention layer;

[0234] A mapping unit, used for mapping the features in the second feature map to a plurality of text characters and probabilities corresponding to the plurality of text characters respectively based on the weights of the features in the second feature map through an activation function in the fully connected layer;

[0235] The second determining unit is used to determine the text character password based on a plurality of text characters and the probabilities respectively corresponding to the plurality of text characters.

[0236] In some embodiments, the second determining module 430 may specifically include:

[0237] A first acquisition submodule, used for acquiring a plurality of training samples before inputting the speech information into the speech recognition model, each training sample including a first speech information sample and its corresponding first character label;

[0238] A second input submodule, used for inputting the first voice information sample into the initial voice recognition model, so as to perform voice recognition on the first voice information sample through the initial voice recognition model to obtain a first predicted character;

[0239] A third determination submodule, used to determine a loss function value according to the similarity between the first predicted character and the first character label;

[0240] The training submodule is used to adjust the first model parameter of the initial speech recognition model based on the back propagation algorithm when the loss function value does not meet the preset training stop condition, and return to execute the first speech information sample input into the initial speech recognition model so as to perform speech recognition on the first speech information sample through the initial speech recognition model to obtain the first predicted character, until the loss function value meets the preset training stop condition, and a trained speech recognition model is obtained.

[0241] In some embodiments, the second determining module 430 may specifically include:

[0242] A second acquisition submodule is used to acquire a plurality of test samples after obtaining the trained speech recognition model, each test sample including a second speech information sample and its corresponding second character label;

[0243] A third input submodule, used for inputting the second voice information sample into the voice recognition model, so as to perform voice recognition on the second voice information sample through the voice recognition model to obtain a second predicted character;

[0244] a fourth determination submodule, configured to determine a character error rate according to similarities between a plurality of second predicted characters and their corresponding first character labels;

[0245] The tuning submodule is used to adjust the second model parameters of the speech recognition model when the character error rate is greater than a preset threshold, and return to execute the second speech information sample input into the speech recognition model so as to perform speech recognition on the second speech information sample through the speech recognition model to obtain a second predicted character until the character error rate is less than or equal to the preset threshold.

[0246] The embodiment of the present application can ensure the security of the correspondence between text characters and numerical characters by displaying multiple text characters and their one-to-one corresponding numerical characters in the password input interface of the wearable smart device, and can also enable the user who inputs the password to clearly understand the encryption information corresponding to the actual numerical character password (i.e., the text character password). Since multiple text characters have different pronunciations, the voice information input by the user is determined in response to the voice input of the user selecting multiple target characters from multiple text characters, the voice information is voice recognized, the text character password is determined, and based on the correspondence between the text characters and the numerical characters, the text character password is converted into a numerical character password, so that the password can be input through voice interaction. On the one hand, the user can enter the password without manually operating the password keyboard, which improves the convenience of password input; on the other hand, since the user who does not wear the wearable smart device cannot know the correspondence between the text characters and the numerical characters, the password security can be guaranteed. In addition, by generating a correspondence between text characters and numeric characters based on system time, target user identification and random numbers, that is, each time in response to a trigger operation for password verification, the displayed virtual keyboard is different, which increases the difficulty of guessing or cracking the password, and further improves the password security. In this way, by receiving the encrypted password information (i.e., the text character password) input by the user's voice, and sending the decrypted numeric character password to the password verification system, so that the password verification system verifies the numeric character password, the password security and the convenience of password input can be guaranteed during the password verification process.

[0247] Based on the password verification method provided in the above embodiment, the embodiment of the present application also provides a specific implementation of the electronic device. Figure 5 A schematic diagram of an electronic device 500 provided in an embodiment of the present application is shown.

[0248] The electronic device 500 may include a processor 510 and a memory 520 storing computer program instructions.

[0249] Specifically, the processor 510 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0250] The memory 520 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 520 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. Where appropriate, the memory 520 may include a removable or non-removable (or fixed) medium. Where appropriate, the memory 520 may be inside or outside the electronic device 500. In a particular embodiment, the memory 520 is a non-volatile solid-state memory.

[0251] The memory may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Thus, typically, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to the first aspect of the present application.

[0252] The processor 510 implements any one of the password verification methods in the above embodiments by reading and executing computer program instructions stored in the memory 520 .

[0253] In one example, the electronic device 500 may further include a communication interface 530 and a bus 540. Figure 5 As shown, the processor 510, the memory 520, and the communication interface 530 are connected via a bus 540 and communicate with each other.

[0254] The communication interface 530 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0255] Bus 540 includes hardware, software or both, and the parts of electronic equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front side bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 540 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the application considers any suitable bus or interconnection.

[0256] Exemplarily, the electronic device 500 may be a mobile phone, a tablet computer, a laptop computer, a PDA, an in-vehicle electronic device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA).

[0257] The electronic device can execute the password verification method in the embodiment of the present application, thereby realizing the combination Figures 2 to 4 Described password verification method and device.

[0258] In addition, in combination with the password verification method in the above embodiment, the embodiment of the present application can provide a computer-readable storage medium for implementation. The computer-readable storage medium stores computer program instructions; when the computer program instructions are executed by a processor, any one of the password verification methods in the above embodiment is implemented.

[0259] In combination with the password verification method in the above embodiment, the embodiment of the present application can provide a computer program product for implementation. When the instructions in the computer program product are executed by a processor of an electronic device, any one of the password verification methods in the above embodiment is implemented.

[0260] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0261] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0262] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0263] The above reference is according to the method of the embodiment of the present application, the flow chart of the device (system) and the computer program product and / or the block diagram described various aspects of the present application.It should be understood that each square box in the flow chart and / or the block diagram and the combination of each square box in the flow chart and / or the block diagram can be realized by computer program instructions.These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the realization of the function / action specified in one or more square boxes of the flow chart and / or the block diagram.Such a processor can be but is not limited to a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit.It can also be understood that each square box in the block diagram and / or the flow chart and the combination of the square boxes in the block diagram and / or the flow chart can also be realized by the dedicated hardware that performs the specified function or action, or can be realized by the combination of dedicated hardware and computer instructions.

[0264] The above is only a specific implementation of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A password verification method, characterized in that: Applied to a wearable smart device, the method comprises: In response to a triggering operation of password verification, a password input interface is displayed, the password input interface includes a virtual keyboard, the virtual keyboard includes a plurality of text characters and their one-to-one corresponding numeric characters, the correspondence between the text characters and the numeric characters is generated based on system time, a target user identifier and a random number, the target user identifier is a user identifier having a binding relationship with the wearable smart device, and the plurality of text characters have different pronunciations; In response to a user's voice input of selecting a plurality of target characters from the plurality of text-type characters, determining voice information input by the user; Performing voice recognition on the voice information to determine a text-based character password; Based on the correspondence between the text characters and the numeric characters, converting the text character password into a numeric character password; The numerical character password is sent to a password verification system so that the password verification system verifies the numerical character password.

2. The method according to claim 1, characterized in that After responding to the triggering operation of password verification and before displaying the password input interface, the method further includes: Acquire a text character set, a plurality of numeric characters, the system time, the target user identifier, and the random number, wherein the text character set includes a plurality of the text characters; Encrypting the plurality of numeric characters based on the system time, the target user identifier and the random number to obtain a plurality of first encrypted information; Acquire the plurality of text characters with different pronunciations respectively corresponding to the plurality of first encrypted information from the text character set; Mapping the plurality of numeric characters to the plurality of text characters to obtain a correspondence between the plurality of numeric characters and the plurality of text characters; The virtual keyboard is determined based on a correspondence between the plurality of numeric characters and the plurality of text characters.

3. The method according to claim 2, characterized in that The step of encrypting the plurality of numeric characters based on the system time, the target user identifier and the random number to obtain a plurality of first encrypted information includes: Generate a target key based on the system time, the target user identifier and the random number through a hash algorithm; Based on the Advanced Encryption Standard algorithm, the target key is used to encrypt the multiple numeric characters to obtain the multiple first encrypted information.

4. The method according to claim 3, characterized in that The converting the text-type character password into a numeric-type character password based on the correspondence between the text-type character and the numeric-type character includes: Based on the correspondence between the text characters and the first encryption information, converting the text character password into an encryption information sequence, wherein the encryption information sequence includes the plurality of first encryption information; Based on the Advanced Encryption Standard algorithm, the encrypted information sequence is decrypted to obtain the numerical character password.

5. The method according to any one of claims 1 to 4, characterized in that: The sending the numerical character password to the password verification system comprises: Encrypting the numerical character password to obtain second encrypted information; The second encrypted information is sent to the password verification system.

6. The method according to any one of claims 1 to 4, characterized in that: In response to a user's voice input of selecting a plurality of target characters from the plurality of text characters, determining the voice information input by the user comprises: authenticating the user and obtaining a verification result; When the verification result is that the verification is passed, receiving a voice input of the user selecting a plurality of target characters from the plurality of text characters; In response to the voice input, voice information of the user input is determined.

7. The method according to claim 1, characterized in that The performing speech recognition on the speech information to determine the text-based character password includes: Inputting the speech information into a speech recognition model; The voice information is subjected to voice recognition through the voice recognition model to determine the text-based character password.

8. The method according to claim 7, characterized in that The speech information is a time domain signal, and the step of inputting the speech information into a speech recognition model includes: Converting the speech information from the time domain signal to a frequency domain signal by fast Fourier transform; Extracting spectral features from the frequency domain signal; The spectral features are input into the speech recognition model.

9. The method according to claim 8, characterized in that The speech recognition model includes a convolution layer, an attention layer, a pooling layer and a fully connected layer; the speech recognition is performed on the speech information by the speech recognition model to determine the text-based character password, including: Performing a convolution operation on the frequency spectrum feature through the convolution layer to obtain a first feature map; Downsampling the first feature map through the pooling layer to obtain a second feature map; Determining weights of features in the second feature map by the attention layer; Mapping the features in the second feature map to a plurality of text characters and probabilities corresponding to the plurality of text characters respectively based on weights of the features in the second feature map through an activation function in the fully connected layer; The text character password is determined based on the plurality of text characters and the probabilities respectively corresponding to the plurality of text characters.

10. The method according to any one of claims 7 to 9, characterized in that: Before inputting the speech information into the speech recognition model, the method further comprises: Acquire multiple training samples, each of which includes a first voice information sample and a first character label corresponding to it; Inputting the first voice information sample into an initial voice recognition model to perform voice recognition on the first voice information sample through the initial voice recognition model to obtain a first predicted character; Determining a loss function value according to the similarity between the first predicted character and the first character label; When the loss function value does not meet the preset training stop condition, the first model parameter of the initial speech recognition model is adjusted based on the back propagation algorithm, and the process of inputting the first speech information sample into the initial speech recognition model is returned to execute, so as to perform speech recognition on the first speech information sample through the initial speech recognition model to obtain a first predicted character, until the loss function value meets the preset training stop condition, thereby obtaining the trained speech recognition model.

11. The method according to claim 10, characterized in that After obtaining the trained speech recognition model, the method further includes: Acquire multiple test samples, each of which includes a second voice information sample and a corresponding second character label; Inputting the second voice information sample into the voice recognition model to perform voice recognition on the second voice information sample through the voice recognition model to obtain a second predicted character; Determining a character error rate based on similarities between a plurality of the second predicted characters and their corresponding first character labels; When the character error rate is greater than a preset threshold, the second model parameters of the speech recognition model are adjusted, and the step of inputting the second speech information sample into the speech recognition model is returned to execute so as to perform speech recognition on the second speech information sample through the speech recognition model to obtain a second predicted character, until the character error rate is less than or equal to the preset threshold.

12. A password verification device, characterized in that: Applied to a wearable smart device, the device comprises: a display module, configured to display a password input interface in response to a trigger operation of password verification, wherein the password input interface includes a virtual keyboard, wherein the virtual keyboard includes a plurality of text characters and corresponding numeric characters thereof, wherein the correspondence between the text characters and the numeric characters is generated based on system time, a target user identifier and a random number, wherein the target user identifier is a user identifier having a binding relationship with the wearable smart device, and wherein the plurality of text characters have different pronunciations; A first determining module, configured to determine the voice information input by the user in response to the user's voice input of selecting a plurality of target characters from the plurality of text-type characters; A second determination module is used to perform voice recognition on the voice information to determine the text-based character password; A conversion module, used for converting the text-type character password into a numeric character password based on the correspondence between the text-type character and the numeric character; The sending module is used to send the numerical character password to the password verification system so that the password verification system verifies the numerical character password.

13. An electronic device, characterized in that: The electronic device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the password verification method as described in any one of claims 1-11 is implemented.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the password verification method according to any one of claims 1 to 11 is implemented.

15. A computer program product, characterized in that When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the password verification method as described in any one of claims 1-11.

Images