Multi-encryption-based AI glasses security payment method and related device

Abstract

The invention discloses a multi-encryption-based AI glasses security payment method and a related device, and relates to the field of security payment. In the method, multi-modal biological information of a user is collected, when the multi-modal biological information is consistent with preset biological information, a dynamic password request is sent to bound target equipment, and a dynamic password returned by the target equipment is received and displayed on a display interface; and when the user inputs the dynamic password on the payment interface, encrypting payment information by using a private key to obtain encrypted information, sending the encrypted information to a payment server to decrypt the payment information through the payment server and verify the decrypted data through a block chain network, and when the verification result is successful, sending the encrypted information to the user. Fund transfer is executed according to the payment information; and fund transfer information returned by the payment server is received. By implementing the technical scheme provided by the invention, the payment security and convenience are improved, and information leakage is effectively prevented through a multi-encryption technology.

Description

Technical Field

[0001] The present application relates to the technical field of secure payment, and specifically to a method and related apparatus for secure payment using AI glasses based on multiple encryption. Background Art

[0002] With the rapid development of artificial intelligence (AI) technology, smart wearable devices are becoming increasingly integrated into people's daily lives. AI glasses, owing to their portability and interactivity, have become a key innovative tool in the payment field. By integrating multiple sensors and computing units, AI glasses enable efficient interaction with users, providing a convenient payment experience and significantly advancing contactless payment technology.

[0003] In existing technologies, single biometric authentication or traditional password verification is often used to ensure payment security. For example, some devices use facial or voice recognition for authentication. However, single biometric authentication methods in existing technologies are vulnerable to counterfeiting attacks.

[0004] Therefore, how to effectively improve the security of identity authentication while ensuring payment convenience has become a technical problem that needs to be solved urgently. Summary of the Invention

[0005] This application provides a method and related device for secure payment using AI glasses based on multiple encryption. By collecting the user's multimodal biometric information (iris image, voice command and gesture image) for identity authentication, and using dynamic password requests and blockchain network verification to complete payment operations, it achieves the effect of improving payment security and convenience, ensuring the user's payment reliability in complex environments, and at the same time effectively preventing information leakage and fraud through multiple encryption technology.

[0006] In a first aspect of the present application, a method for secure payment using AI glasses based on multiple encryption is provided, which is applied to AI glasses. The method comprises: In response to a user's operation of activating a payment function, collecting multimodal biometric information of the user and determining whether the multimodal biometric information is consistent with preset biometric information, the multimodal biometric information including an iris image, a voice command, and a gesture image; When the multimodal biometric information is consistent with the preset biometric information, a dynamic password request is sent to the bound target device, and the dynamic password returned by the target device is received and displayed on the display interface; When the user enters the dynamic password on the payment interface, the payment information is encrypted using the private key to obtain encrypted information, and the encrypted information is sent to the payment server for decryption by the payment server and verification of the decrypted data through the blockchain network. When the verification result is passed, the funds are transferred according to the payment information; Receive the funds transfer information returned by the payment server.

[0007] Optionally, collecting multimodal biometric information of the user and determining whether the multimodal biometric information is consistent with preset biometric information includes: Capturing an iris image through a camera, processing the iris image based on a lightweight convolutional neural network to obtain an encrypted feature vector, and calculating the cosine similarity between the encrypted feature vector and a preset feature vector; Collecting a voice signal through a microphone, processing the voice signal based on a Mel-frequency cepstral coefficient algorithm to obtain multidimensional feature parameters, and calculating a voiceprint feature dynamic time warping score based on the multidimensional feature parameters and a preset voiceprint template; Capturing gesture trajectories through a dynamic visual sensor, processing the gesture trajectories based on a long short-term memory network to obtain a spatiotemporal coding sequence, and calculating the gesture trajectory edit distance based on the spatiotemporal coding sequence and a preset gesture template; It is determined whether the cosine similarity, the voiceprint feature dynamic time warping score, and the gesture trajectory edit distance all meet preset conditions.

[0008] Optionally, the processing of the speech signal based on a Mel-frequency cepstral coefficient algorithm to obtain multidimensional feature parameters, and calculating the voiceprint feature dynamic time warping score according to the multidimensional feature parameters and a preset voiceprint template include: Performing a fast Fourier transform and a discrete cosine transform on each frame of the speech signal to obtain a basic Mel-frequency cepstral coefficient feature, calculating a first-order difference and a second-order difference of the basic Mel-frequency cepstral coefficient feature, and expanding the basic Mel-frequency cepstral coefficient feature into a multidimensional feature parameter according to the first-order difference and the second-order difference; An MFCC feature sequence is constructed based on the multidimensional feature parameters corresponding to all frame signals of the speech signal, and the Mel-frequency cepstral coefficient feature sequence is compared with a preset voiceprint template to construct a comparison matrix. The optimal path from the upper left corner to the lower right corner in the comparison matrix is determined through dynamic programming. The distances of all matching frames are accumulated according to the optimal path to obtain a total cumulative distance. The total cumulative distance is divided by the number of all matching frames to obtain the dynamic time warping score of the voiceprint feature.

[0009] Optionally, the processing the gesture trajectory based on the long short-term memory network to obtain a spatiotemporal coding sequence, and calculating the gesture trajectory edit distance according to the spatiotemporal coding sequence and a preset gesture template includes: Processing the spatial coordinates of the target finger in each frame image of the gesture trajectory to obtain a spatiotemporal coding vector, and constructing a spatiotemporal coding sequence based on all the spatiotemporal coding vectors of the gesture trajectory; A gesture feature dynamic time warping score is calculated based on the spatiotemporal coding sequence and the preset gesture template, and a gesture trajectory edit distance is determined based on the gesture feature dynamic time warping score.

[0010] Optionally, sending a dynamic password request to a bound target device, and receiving and displaying the dynamic password returned by the target device on a display interface includes: Sending a dynamic password request to the bound target device through an elliptic curve encryption channel based on the device fingerprint, wherein the device fingerprint includes the device identification of the AI glasses; Receive the return data of the target device through the main channel, decrypt the return data to obtain the password plaintext, and display the dynamic password in plaintext on the display interface; A voice verification code is obtained through a backup channel, where the voice verification code is consistent with the dynamic password.

[0011] Optionally, sending a dynamic password request to a bound target device through an elliptic curve encryption channel according to the device fingerprint includes: Determine a base point of the elliptic curve, generate a first random number as a private key of the elliptic curve, and generate a public key based on the private key and the base point. Binary concatenation of the device identifier and the private key to obtain concatenated data, performing a hash operation on the concatenated data to generate a hash value, and encoding the hash value to obtain a device fingerprint; A password request is generated according to the device fingerprint, the transaction information, and the second random number, and the password request is encrypted using the public key of the target device to obtain a dynamic password request.

[0012] Optionally, sending the encrypted information to a payment server for decryption by the payment server and verifying the decrypted data through a blockchain network includes: The encrypted information is decrypted using a public key algorithm in asymmetric encryption technology to obtain the payment information, the payment information is formatted according to the protocol requirements of the blockchain network to obtain a payment request, and the payment request is broadcast to multiple nodes in the blockchain network. The number of nodes that recognize the transaction legitimacy of the payment request is determined, and a ratio of the number to the total number of nodes is calculated. When the ratio exceeds a threshold, the payment request is confirmed to be legal.

[0013] In a second aspect of the present application, a system for secure payment using AI glasses based on multiple encryption is provided, comprising a collection module, a request module, a sending module, and a receiving module, wherein: a collection module configured to collect multimodal biometric information of the user in response to an operation by the user to activate a payment function, and determine whether the multimodal biometric information is consistent with preset biometric information, the multimodal biometric information including an iris image, a voice command, and a gesture image; a request module configured to send a dynamic password request to a bound target device when the multimodal biometric information is consistent with the preset biometric information, and receive and display the dynamic password returned by the target device on a display interface; a sending module configured to, when the user enters the dynamic password on the payment interface, encrypt the payment information using a private key to obtain encrypted information, and send the encrypted information to the payment server for decryption by the payment server and verification of the decrypted data through the blockchain network; when the verification result is passed, execute the fund transfer according to the payment information; The receiving module is configured to receive the funds transfer information returned by the payment server.

[0014] In the third aspect of the present application, an electronic device is provided, including a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device performs any of the methods described above.

[0015] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores instructions. When the instructions are executed, any one of the methods described above is executed.

[0016] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. By collecting multimodal biometric information from users, including iris images, voice commands, and gesture images, the accuracy and security of identity verification are greatly improved compared to single biometric identification methods. The combination of multiple biometric features makes it extremely difficult to forge or impersonate others to make payments, effectively preventing identity theft and fraud, and providing more reliable protection for user funds. 2. After the biometric information is verified, a dynamic password request is sent to the bound target device, and the dynamic password is displayed upon receipt. The dynamic password is time-sensitive and one-time. A different password is used for each payment. Even if the password is intercepted during transmission, it cannot be used again, further reducing the risk of payment information being stolen and abused, and improving the security of the payment process. 3. Encrypt the payment information using the private key. Only the payment server with the corresponding public key can decrypt the information. This encryption method ensures the confidentiality of payment information during transmission, preventing man-in-the-middle attacks and data leaks. Even if the data is intercepted during transmission, attackers cannot obtain sensitive payment information. 4. The decrypted data is verified through the blockchain network. The distributed ledger nature of blockchain makes data immutable once recorded, ensuring its integrity and credibility. Blockchain verification effectively prevents payment information from being tampered with or forged, ensuring the authenticity and legitimacy of payment transactions and providing an additional layer of security for the payment process. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flowchart of a method for secure payment using AI glasses based on multiple encryption, disclosed in an embodiment of the present application; Figure 2 This is a module diagram of a system for secure payment using AI glasses based on multiple encryption disclosed in an embodiment of the present application; Figure 3 This is a structural diagram of an electronic device disclosed in an embodiment of the present application.

[0018] Explanation of the reference numerals: 201, acquisition module; 202, request module; 203, sending module; 204, receiving module; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. DETAILED DESCRIPTION

[0019] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0020] In the description of the embodiments of this application, words such as "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "for example" or "for instance" in the embodiments of this application should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "for example" or "for instance" is intended to present the relevant concepts in a concrete manner.

[0021] In the description of the embodiments of the present application, the term "multiple" means two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0022] This embodiment discloses a method for secure payment on AI glasses based on multiple encryption, which is applied to AI glasses. Figure 1 This is a flow chart of the method for secure payment using AI glasses based on multiple encryption disclosed in an embodiment of the present application, such as Figure 1 As shown, the method includes the following steps: S101. In response to a user initiating a payment function, collecting multimodal biometric information of the user and determining whether the multimodal biometric information is consistent with preset biometric information, the multimodal biometric information including an iris image, a voice command, and a gesture image; S102: When the multimodal biometric information is consistent with the preset biometric information, a dynamic password request is sent to the bound target device, and the dynamic password returned by the target device is received and displayed on a display interface; S103. When the user enters the dynamic password on the payment interface, the payment information is encrypted using the private key to obtain encrypted information, and the encrypted information is sent to the payment server for decryption by the payment server. The decrypted data is verified through the blockchain network. When the verification result is passed, the funds are transferred according to the payment information; S104: Receive the funds transfer information returned by the payment server.

[0023] Users can initiate payment through specific voice commands, gestures, or physical buttons on the glasses. The AI glasses are equipped with a dedicated iris recognition camera that quickly and accurately captures the user's iris image. The iris is highly unique and stable, and each person's iris texture is unique, making it a highly reliable biometric feature. The glasses' built-in microphone captures specific voice commands from the user. These commands can be preset payment passphrases or specific wake-up words. The system analyzes and extracts voice characteristics, such as pitch, speed, and pronunciation. The camera on the glasses captures user gestures, such as specific waving gestures or finger combinations. The system processes and analyzes the gesture images to identify gesture characteristics and patterns. The collected multimodal biometric information is compared with the preset biometric information stored in the system. The preset biometric information is collected and encrypted when the user registers for the payment function. Only when the captured iris image, voice command, and gesture image all match the preset biometric information is the identity verification considered successful and the next step is taken. Once multimodal biometric verification is successful, the AI glasses will send a dynamic password request to the bound target device (such as the user's smartphone). This request is typically transmitted via a secure communication protocol to ensure information security. Upon receiving the request, the target device generates a dynamic password and returns it to the AI glasses via a secure channel. The AI glasses will clearly display the dynamic password on the user's display interface, allowing them to enter it on the payment screen. Dynamic passwords have a time limit and typically expire automatically after a certain period of time, enhancing payment security. After the user enters the dynamic password on the payment screen, the AI glasses will encrypt the payment information using a pre-stored private key. The payment information may include sensitive data such as the payment amount and the recipient's account number. The private key is a unique encryption key owned by the user. Only the corresponding public key can decrypt the information encrypted with the private key, ensuring the confidentiality of the payment information during transmission. The encrypted payment information (encrypted information) is then sent to the payment server. The payment server decrypts the encrypted information using the corresponding public key to obtain the original payment information. The payment server verifies the decrypted payment information through the blockchain network. The blockchain network is decentralized, immutable, and traceable, enabling verification of the authenticity, integrity, and legitimacy of payment information. This verification process may involve checking transaction signatures and verifying the source and destination of transactions. If the blockchain network verifies that the payment information is successful, the payment server will execute a funds transfer based on the payment information, transferring the corresponding funds from the user's account to the recipient's account. The AI glasses receive the funds transfer information returned by the payment server. This information may include details such as whether the transfer was successful, the transfer time, and the transfer amount. Users can view the funds transfer information on the AI glasses' display interface to understand the payment results.At the same time, the system can also store this information locally or upload it to the cloud so that users can subsequently query and manage payment records.

[0024] Optionally, collecting multimodal biometric information of the user and determining whether the multimodal biometric information is consistent with preset biometric information includes: Capturing an iris image through a camera, processing the iris image based on a lightweight convolutional neural network to obtain an encrypted feature vector, and calculating the cosine similarity between the encrypted feature vector and a preset feature vector; Collecting a voice signal through a microphone, processing the voice signal based on a Mel-frequency cepstral coefficient algorithm to obtain multidimensional feature parameters, and calculating a voiceprint feature dynamic time warping score based on the multidimensional feature parameters and a preset voiceprint template; Capturing gesture trajectories through a dynamic visual sensor, processing the gesture trajectories based on a long short-term memory network to obtain a spatiotemporal coding sequence, and calculating the gesture trajectory edit distance based on the spatiotemporal coding sequence and a preset gesture template; It is determined whether the cosine similarity, the voiceprint feature dynamic time warping score, and the gesture trajectory edit distance all meet preset conditions.

[0025] The AI glasses' high-precision camera captures the user's iris image. This camera must have high resolution, fast focus, and excellent optical performance to ensure clear and accurate capture of iris details. The captured iris image is processed using a lightweight convolutional neural network (CNN). A lightweight CNN is a specially designed neural network architecture that maintains a certain level of accuracy while maintaining a small model size and low computational complexity, making it suitable for resource-constrained AI glasses. This network extracts a representative encrypted feature vector from the iris image. This feature vector contains the iris' unique texture information. The cosine similarity between the extracted encrypted feature vector and the preset feature vector is calculated. Cosine similarity measures the similarity between two vectors, ranging from -1 to 1. Values closer to 1 indicate greater similarity. The preset feature vector is the iris feature vector collected during user registration and processed in the same way. By comparing the cosine similarity, the consistency between the currently captured iris image and the preset iris information can be determined. The user's voice signal is collected using the built-in microphone of the AI glasses. The microphone must have high sensitivity and low noise to accurately capture the user's voice commands. The collected voice signal is processed using the Mel-Frequency Cepstral Coefficient (MFCC) algorithm. MFCC is a feature extraction method widely used in speech recognition. It converts voice signals into multidimensional feature parameters that reflect the spectral characteristics of speech, such as pitch and timbre. The dynamic time warping (DTW) score of the voiceprint feature is calculated based on the extracted multidimensional feature parameters and the preset voiceprint template. The preset voiceprint template is the voice feature template collected during user registration. DTW is an algorithm used to measure the similarity between two time series. It can handle the potential stretching and shifting of voice signals on the time axis. The DTW score is calculated to assess the degree of match between the currently collected voice signal and the preset voiceprint template. The user's gesture trajectory is captured using a dynamic vision sensor (DVS). The DVS is an event-driven vision sensor that records gesture motion information with extremely high temporal resolution and can accurately capture rapidly changing gestures. The captured gesture trajectory is processed using a long short-term memory (LSTM) network. LSTM is a special recurrent neural network that effectively handles long-term dependencies in sequential data. Using the LSTM network, gesture trajectories can be converted into spatiotemporal encoding sequences that contain spatiotemporal features such as the gesture's direction, speed, and amplitude. The edit distance of the gesture trajectory is calculated based on the extracted spatiotemporal encoding sequence and a preset gesture template. A preset gesture template is a predefined set of standard gesture trajectory features. The edit distance measures the difference between two sequences. By calculating the edit distance, we can determine the degree of similarity between the currently captured gesture trajectory and the preset gesture template.Determine whether the calculated cosine similarity, voiceprint dynamic time warping score, and gesture trajectory edit distance all meet preset conditions. Preset conditions are thresholds or ranges set based on actual application scenarios and security requirements. For example, cosine similarity must be greater than a certain threshold, voiceprint dynamic time warping score must be less than a certain threshold, and gesture trajectory edit distance must be within a certain range. Only when all three indicators meet the preset conditions is the collected multimodal biometric information considered consistent with the preset biometric information, and thus identity verification is passed.

[0026] Lightweight CNNs can reduce computational complexity and increase processing speed while maintaining a certain level of recognition accuracy, enabling rapid iris image recognition. Cosine similarity calculation accurately measures the similarity between the collected iris feature vector and the preset feature vector, effectively determining whether they represent the same person and improving identification accuracy. The use of encrypted feature vectors enhances data security. Even if the feature vector is intercepted during transmission or storage, it is difficult to decrypt and restore it to the original iris image, thus protecting the user's biometric privacy. The MFCC algorithm extracts key features from speech signals, and its multidimensional feature parameters comprehensively reflect the acoustic characteristics of speech. The DTW score accounts for the dynamic changes in speech signals over time, accurately matching voiceprint features at varying speech rates and intonations, improving the robustness and accuracy of voiceprint recognition. This method is highly adaptable to diverse environments and speaking styles, effectively recognizing voiceprints even in the presence of slight background noise or when the user's pronunciation habits vary, ensuring reliable authentication. Dynamic vision sensors can accurately capture gesture trajectories in real time. LSTM networks excel at processing sequential data, effectively extracting spatiotemporal features from gesture trajectories and effectively encoding and recognizing complex gestures. The edit distance calculation for gesture trajectories accounts for subtle differences in gesture trajectories, making gesture verification more flexible and accurate. Users can complete authentication through natural gestures, making interaction more convenient and engaging.

[0027] Optionally, the processing of the speech signal based on a Mel-frequency cepstral coefficient algorithm to obtain multidimensional feature parameters, and calculating the voiceprint feature dynamic time warping score according to the multidimensional feature parameters and a preset voiceprint template include: Performing a fast Fourier transform and a discrete cosine transform on each frame of the speech signal to obtain a basic Mel-frequency cepstral coefficient feature, calculating a first-order difference and a second-order difference of the basic Mel-frequency cepstral coefficient feature, and expanding the basic Mel-frequency cepstral coefficient feature into a multidimensional feature parameter according to the first-order difference and the second-order difference; An MFCC feature sequence is constructed based on the multidimensional feature parameters corresponding to all frame signals of the speech signal, and the Mel-frequency cepstral coefficient feature sequence is compared with a preset voiceprint template to construct a comparison matrix. The optimal path from the upper left corner to the lower right corner in the comparison matrix is determined through dynamic programming. The distances of all matching frames are accumulated according to the optimal path to obtain a total cumulative distance. The total cumulative distance is divided by the number of all matching frames to obtain the dynamic time warping score of the voiceprint feature.

[0028] Fast Fourier Transform (FFT): Perform an FFT on each frame of the speech signal to convert the time domain signal into the frequency domain. This is because frequency domain information better reflects the acoustic characteristics of speech signals, and different speech patterns have unique distribution patterns in the frequency domain. Discrete Cosine Transform (DCT): Perform a DCT on the frequency domain signal after the FFT to obtain basic Mel-frequency cepstral coefficient features. The DCT converts the signal from the frequency domain to the cepstral domain, further compressing information, highlighting the signal's key features, and reducing data dimensionality and redundancy. The first-order and second-order differences of the basic MFCC features are calculated. The first-order differences reflect the rate of change of the feature over time, while the second-order differences reflect the trend of the rate of change. By incorporating differential information, the dynamic characteristics of the speech signal can be captured, as dynamic changes in speech (such as rising and falling pitch) are also important features for voiceprint recognition. The basic MFCC features are combined with their first-order and second-order differences to form multidimensional feature parameters. These multidimensional feature parameters comprehensively describe the static and dynamic characteristics of the speech signal, providing richer information for subsequent voiceprint recognition. A MFCC feature sequence is constructed based on the multidimensional feature parameters corresponding to all frames of the speech signal. This sequence represents the characteristic variations of the entire speech signal along the time axis, with each frame's feature parameter corresponding to an element in the sequence. The constructed MFCC feature sequence is compared with a preset voiceprint template to construct a comparison matrix. The preset voiceprint template is a standard voiceprint feature sequence extracted and stored during user registration. Each element in the comparison matrix represents the feature distance between the two sequences in corresponding frames, typically calculated using methods such as Euclidean distance. A dynamic programming algorithm is used to determine the optimal path from the upper left corner to the lower right corner of the comparison matrix. The core idea of dynamic programming is to decompose the problem into subproblems and save the solutions to these subproblems to avoid repeated computation. In DTW, the minimum cumulative distance from the starting point to each point is gradually calculated to ultimately find the optimal path from the upper left corner to the lower right corner that minimizes the matching cost between the two sequences. The distances of all matching frames along the optimal path are accumulated to obtain the total cumulative distance. This total cumulative distance reflects the overall difference between the two sequences. The total cumulative distance is divided by the number of matching frames to obtain the dynamic time warping score of the voiceprint feature. This score is a normalized value used to measure the similarity between two voiceprint sequences. The lower the score, the more similar the two sequences are, that is, the higher the match between the speech signal and the preset voiceprint template.

[0029] The FFT transforms the time-domain signal into the frequency domain, clearly displaying the frequency distribution of the speech signal. The DCT further processes the frequency domain signal to extract Mel-frequency cepstral coefficients, which reflect the characteristics of the speech vocal tract. These coefficients effectively capture the spectral envelope of speech and are important features of the speech signal. Computing first-order and second-order differences captures the changing trend and acceleration of features, expanding the basic features into multidimensional feature parameters. This enriches the feature dimensionality and makes the features more robust to subtle variations and noise in the speech signal. For example, under varying speaking speeds, intonations, or ambient noise, the multidimensional feature parameters can effectively preserve the essential characteristics of speech, improving the accuracy of voiceprint recognition. Constructing an MFCC feature sequence fully represents the characteristic variations of the speech signal across the entire timeline, providing a comprehensive data foundation for subsequent comparison with a pre-set voiceprint template. Constructing a comparison matrix quantifies the differences between the two feature sequences, making the subsequent matching process more intuitive and accurate. The comparison matrix clearly demonstrates the similarity between the features of each frame, providing effective input for the dynamic time warping algorithm. The Dynamic Time Warping algorithm can handle feature sequences of varying lengths. By finding the optimal path, it can effectively match speech signals at varying speech rates. Even if the speaker's speaking rate differs from that in the preset voiceprint template, it can accurately find a corresponding match, improving the adaptability of voiceprint recognition. The total cumulative distance reflects the overall difference between the two feature sequences. Dividing this by the number of matching frames yields a Dynamic Time Warping score, making the score comparable and intuitive. The lower the score, the more similar the two feature sequences are. This allows for an accurate assessment of the match between the speech signal and the preset voiceprint template, providing a reliable basis for identity verification.

[0030] Optionally, the processing the gesture trajectory based on the long short-term memory network to obtain a spatiotemporal coding sequence, and calculating the gesture trajectory edit distance according to the spatiotemporal coding sequence and a preset gesture template includes: Processing the spatial coordinates of the target finger in each frame image of the gesture trajectory to obtain a spatiotemporal coding vector, and constructing a spatiotemporal coding sequence based on all the spatiotemporal coding vectors of the gesture trajectory; A gesture feature dynamic time warping score is calculated based on the spatiotemporal coding sequence and the preset gesture template, and a gesture trajectory edit distance is determined based on the gesture feature dynamic time warping score.

[0031] For each frame of a gesture trajectory, the system accurately locates and extracts the spatial coordinates of the target finger (such as the index finger or thumb, used for specific gestures). These coordinates reflect the finger's position in three-dimensional space and serve as the fundamental data for describing the gesture trajectory. The acquired spatial coordinates are processed using an LSTM network. LSTM excels at processing sequential data, capturing the temporal evolution of the finger's spatial coordinates and mining their temporal features and spatial correlations. By analyzing and learning from the coordinate data, the LSTM converts the spatial coordinates in each frame into a spatiotemporal encoding vector. This vector not only contains the finger's current position but also incorporates the motion trends and features from previous moments, providing a more comprehensive representation of the dynamic changes in the gesture. The spatiotemporal encoding vectors corresponding to all frames in the gesture trajectory are combined in chronological order to form a spatiotemporal encoding sequence. This sequence fully records the finger's motion trajectory and dynamic features from the start to the end of the gesture, providing a rich data foundation for subsequent comparison with pre-set gesture templates. The constructed spatiotemporal encoding sequence is then compared with the pre-set gesture templates. Preset gesture templates are standard gesture feature sequences pre-stored during the system registration or training phase, representing legal or specific gesture patterns. The dynamic time warping (DTW) algorithm is used to calculate the similarity between two sequences. The DTW algorithm can handle sequences of varying lengths and calculates the cumulative distance between the two sequences by finding the optimal temporal alignment path. During this process, the algorithm takes into account the scaling and offset of the gesture trajectory along the time axis, ensuring an accurate measure of the similarity between the two gestures even when the gesture execution speed varies. The resulting cumulative distance, after certain processing (such as normalization), is converted into the gesture feature dynamic time warping score. A lower score indicates greater similarity between the two gesture sequences. The gesture trajectory edit distance is determined based on the gesture feature dynamic time warping score. It can be understood that the gesture feature dynamic time warping score, to a certain extent, reflects the minimum cost of operations (such as insertion, deletion, and substitution) required to transform the spatiotemporal encoding sequence into the preset gesture template. Therefore, the gesture feature dynamic time warping score can be directly used as the gesture trajectory edit distance, or further transformations and calculations based on the score can be performed to more intuitively represent the degree of difference between two gesture trajectories. The smaller the gesture trajectory edit distance is, the more closely the current gesture trajectory matches the preset gesture template.

[0032] LSTM networks excel at processing sequential data and can effectively process the spatial coordinates of the target finger in each frame of a gesture trajectory. By extracting information such as the coordinate's changing trend and speed and converting it into a spatiotemporal encoding vector, they accurately capture the dynamic characteristics of the gesture in both time and space. For example, when recognizing a hand gesture, the LSTM can analyze the finger's position and direction at different moments, accurately describing the gesture's trajectory. Constructing a spatiotemporal encoding sequence integrates the spatiotemporal encoding vectors of all frames, fully representing the characteristics of the entire gesture trajectory. This comprehensive representation provides a rich information foundation for subsequent comparison with pre-set gesture templates, enabling gesture recognition to consider the entire gesture process and its detailed variations. The Dynamic Time Warping (DTW) algorithm can handle spatiotemporal encoding sequences of varying lengths. By calculating the dynamic time warping score of gesture features, it can effectively match gesture trajectories with different rhythms and speeds. For example, users may perform gestures at varying speeds. The DTW algorithm can find the optimal matching path between two sequences, accurately measuring the similarity between them even when the rhythm of the gestures differs from the pre-set template. Accurately quantifying gesture differences: The edit distance of a gesture trajectory is determined by the dynamic time warping score, which quantifies the difference between the spatiotemporal encoding sequence and the preset gesture template. The smaller the edit distance, the more similar the two gesture trajectories are, allowing accurate judgment of whether the user's gesture matches the preset gesture template. This quantification method makes gesture recognition results more intuitive and reliable.

[0033] Optionally, sending a dynamic password request to a bound target device, and receiving and displaying the dynamic password returned by the target device on a display interface includes: Sending a dynamic password request to the bound target device through an elliptic curve encryption channel based on the device fingerprint, wherein the device fingerprint includes the device identification of the AI glasses; Receive the return data of the target device through the main channel, decrypt the return data to obtain the password plaintext, and display the dynamic password in plaintext on the display interface; A voice verification code is obtained through a backup channel, where the voice verification code is consistent with the dynamic password.

[0034] The device fingerprint contains the device identifier of the AI glasses, which is a unique identifier similar to a person's ID number. It can be a hardware serial number assigned during the device's manufacturing process or a unique identifier generated at the software level. Elliptic curve cryptography is used to establish an encrypted channel. Elliptic curve cryptography is a public-key encryption algorithm based on elliptic curve mathematical problems, offering advantages such as high security, short key length, and high computational efficiency. Sending dynamic password requests over this encrypted channel ensures the confidentiality and integrity of the request information during transmission, preventing theft or tampering. For example, when sending a request, the AI glasses encrypt the request data using the public key of the target device (such as a mobile phone). Only the target device can decrypt the request using its private key. The primary channel is the primary channel for data transmission between the AI glasses and the target device. It has high bandwidth and stability, enabling fast and accurate data transmission. After receiving the dynamic password request, the target device sends the return data containing the dynamic password to the AI glasses via the primary channel. Upon receiving the return data, the AI glasses decrypt the encrypted data using a pre-stored key. This key is negotiated and determined when the AI glasses and the target device are bound and is known only to both parties. After decryption, the AI glasses can obtain the plaintext password, that is, the specific content of the dynamic password. The AI glasses will display the dynamic password in plain text on their display interface for easy viewing and use. The display interface can be a small display screen on the glasses or an external display device connected to the glasses. The backup channel is set up to ensure that the dynamic password can be obtained normally in the event of a failure or interference with the primary channel. It can be a different communication method, such as Bluetooth or Wi-Fi Direct. The voice verification code obtained through the backup channel is the same as the dynamic password obtained through the primary channel. When the target device sends back data, it will also generate a voice verification code and send it to the AI glasses via the backup channel. The user can obtain the verification code by listening to the voice prompt and compare it with the dynamic password on the display interface to further ensure the accuracy and security of the dynamic password. For example, if the user suspects that the dynamic password on the display interface may have been tampered with, they can verify it by listening to the voice verification code.

[0035] Elliptic curve cryptography is highly secure, providing strong encryption with limited key lengths. Using an elliptic curve cryptographic channel to transmit dynamic password requests effectively prevents requests from being eavesdropped or tampered with during transmission, ensuring the confidentiality and integrity of communications. The AI glasses device identifier in the device fingerprint uniquely identifies the device sending the request, ensuring that only legitimate AI glasses can send dynamic password requests to the target device, preventing malicious attacks and forged requests from unauthorized devices. The primary channel, as the primary data transmission path, ensures stable transmission of returned data. Decrypting the returned data accurately retrieves the plaintext dynamic password generated by the target device, ensuring the accuracy of the password information. Displaying the dynamic password in plaintext on the display interface facilitates direct user viewing and use, improving the user experience. This intuitive display also facilitates password verification, avoiding input errors caused by unclear display. The backup channel provides a redundant mechanism for obtaining dynamic passwords. If the primary channel fails or is disrupted, the backup channel ensures that users can still obtain a valid dynamic password, improving system reliability and stability. The voice verification code is consistent with the dynamic password, allowing users to confirm their password information by listening to the voice verification code. This multi-mode verification mechanism increases the accuracy of users obtaining passwords, while also providing users with more choices and meeting the usage habits of different users.

[0036] Optionally, sending a dynamic password request to a bound target device through an elliptic curve encryption channel according to the device fingerprint includes: Determine a base point of the elliptic curve, generate a first random number as a private key of the elliptic curve, and generate a public key based on the private key and the base point. Binary concatenation of the device identifier and the private key to obtain concatenated data, performing a hash operation on the concatenated data to generate a hash value, and encoding the hash value to obtain a device fingerprint; A password request is generated according to the device fingerprint, the transaction information, and the second random number, and the password request is encrypted using the public key of the target device to obtain a dynamic password request.

[0037] Select a suitable elliptic curve and determine its base point G. Elliptic curve cryptography is based on the discrete logarithm problem on elliptic curves. Base point G is a fixed point on the elliptic curve and is used to generate key pairs. Generate a first random number d as the private key for the elliptic curve. The private key is a large, random integer that must be kept strictly confidential. It serves as the basis for generating the public key and performing encryption and signing operations. Based on the private key d and the base point G, the public key Q is generated through the point multiplication Q = dG on the elliptic curve. The public key can be made public and used to encrypt data or verify signatures. Binary concatenate the device identifier (the unique identifier of the AI glasses) and the generated private key d to generate the concatenated data. The device identifier uniquely identifies the device sending the request. Concatenating it with the private key increases the uniqueness and security of the device fingerprint. Perform a hash operation on the concatenated data to generate a hash value. Hash functions are one-way and collision-resistant, capable of converting input data of any length into a fixed-length output value. Different input data are unlikely to generate the same hash value. Encode the hash value to obtain the device fingerprint. The purpose of encoding is to convert the hash value into a format suitable for transmission and storage for subsequent use. A password request is generated based on the device fingerprint, transaction information, and a second random number. Transaction information may include the payment amount and the recipient's account number. The second random number is used to increase the randomness and unpredictability of the request and prevent replay attacks. The device fingerprint is used to identify the device sending the request, ensuring its legitimacy. The password request is encrypted using the target device's public key to generate a dynamic password request. This encryption process uses elliptic curve cryptography. Only the recipient with the target device's private key can decrypt the request, ensuring the confidentiality and integrity of the request during transmission.

[0038] Elliptic curve cryptography (ECC) is based on the intractability of the elliptic curve discrete logarithm problem. Given the same key length, ECC offers higher security than traditional algorithms like RSA. Determining the appropriate elliptic curve base point is fundamental to building a secure encryption system. The resulting private and public key pairs provide key security for subsequent encrypted communications. This key generation method, which generates a first random number as the private key and rapidly generates the public key based on elliptic curve mathematical operations, is relatively simple and efficient. It ensures security while reducing computing resource consumption and improving system efficiency. The device ID is the unique identifier of the AI glasses. This ID is binary concatenated with the private key to generate concatenated data, which is then hashed and encoded to create the device fingerprint. Due to the randomness and uniqueness of the private key, each device's fingerprint is unique, effectively distinguishing one device from another and preventing device impersonation or counterfeiting. Hashing has an avalanche effect, meaning that even small changes in the input data can result in significant differences in the output hash value. By hashing the concatenated data to generate a hash value, the integrity of the device ID and private key is ensured. Any tampering with the data during transmission or storage will cause the hash value to change, allowing for timely detection of data anomalies. The device fingerprint, transaction information, and a second random number are combined to generate a password request. The device fingerprint is used for device authentication, the transaction information clarifies the purpose of the request and related business data, and the second random number increases the randomness and unpredictability of the request, ensuring the integrity and accuracy of the password request information. The password request is encrypted using the target device's public key. Only the target device with the corresponding private key can decrypt the request, effectively preventing the password request from being stolen or tampered with during transmission. This encryption method ensures the secure communication between the AI glasses and the target device, ensuring that the dynamic password request can be securely transmitted to the target device.

[0039] Optionally, sending the encrypted information to a payment server for decryption by the payment server and verifying the decrypted data through a blockchain network includes: The encrypted information is decrypted using a public key algorithm in asymmetric encryption technology to obtain the payment information, the payment information is formatted according to the protocol requirements of the blockchain network to obtain a payment request, and the payment request is broadcast to multiple nodes in the blockchain network. The number of nodes that recognize the transaction legitimacy of the payment request is determined, and a ratio of the number to the total number of nodes is calculated. When the ratio exceeds a threshold, the payment request is confirmed to be legal.

[0040] Asymmetric encryption uses a pair of keys: a public key and a private key. The public key is publicly available and used to encrypt information; the private key is kept secret by the recipient and used to decrypt the information. In this scenario, the payment server decrypts the encrypted information using the public key agreed upon with the sender. This ciphertext is converted into plaintext, resulting in the payment information. Payment information typically includes key information such as the transaction amount, the identities of the two parties, and the transaction time. The payment information is formatted according to the blockchain network's protocol to generate a payment request. Blockchain networks have specific protocols and data format requirements to ensure that transactions can be correctly identified and processed between different nodes. The payment server, in accordance with the blockchain network's protocol specifications, fills the decrypted payment information into the corresponding fields, forming a payment request that complies with the blockchain network's requirements. This ensures that the payment request can be correctly parsed and processed within the blockchain network. The payment request is then broadcast to multiple nodes within the blockchain network. A blockchain network consists of multiple geographically distributed nodes that collectively maintain the blockchain ledger. Each node maintains a complete copy of the blockchain data and is responsible for verifying and recording new transactions. The payment server sends the formatted payment request to multiple nodes within the blockchain network. Broadcasting can involve sending payment requests via network protocols to nodes directly connected to the payment server. These nodes then forward the payment requests to other nodes, ultimately ensuring that all nodes in the blockchain network receive the payment request. The number of nodes that approve the transaction as legitimate is determined. The ratio of this number to the total number of nodes is calculated. When this ratio exceeds a threshold, the payment request is deemed legitimate. Upon receiving a payment request, each node in the blockchain network verifies it according to pre-set rules and algorithms. Verification includes verifying the validity of the transaction signature, the reasonableness of the transaction amount, and sufficient balances between both parties. The ratio of the number of nodes approving the payment request to the total number of nodes is calculated to determine whether the payment request has been approved by a majority of nodes. When this ratio exceeds a pre-set threshold, it indicates that a majority of nodes consider the payment request legitimate, thus confirming the validity of the payment request. This consensus mechanism ensures the security and reliability of the blockchain network and prevents attacks and fraud by malicious nodes.

[0041] In asymmetric encryption, a public key is used for encryption and a private key is used for decryption. Only the payment server possesses the private key corresponding to the public key and can correctly decrypt the encrypted information. This ensures that even if payment information is intercepted during transmission, attackers cannot obtain sensitive information, thus guaranteeing the security of payment information. Decryption capabilities are restricted to specific payment servers, preventing unauthorized entities from decrypting encrypted information and further enhancing the security of the payment system. Blockchain networks have specific protocols and data format requirements. Formatting payment information according to these requirements ensures that payment requests conform to the blockchain network's processing specifications, ensuring correct data transmission and processing within the network and avoiding data processing errors caused by format incompatibility. Formatted payment requests facilitate rapid parsing and processing by nodes in the blockchain network, improving the efficiency of the entire payment verification process. Broadcasting payment requests to multiple nodes allows all nodes in the blockchain network to participate in the verification process. This distributed verification mechanism increases the reliability and fairness of verification and avoids errors or fraud that may occur when a single node verifies. Even if some nodes fail or are attacked, other functioning nodes can continue to verify the payment request, ensuring the stable operation of the entire payment system. A typical consensus mechanism uses a ratio of the number of nodes that approve a payment request to the total number of nodes, and then determines the legitimacy of a payment request based on a set threshold. This mechanism ensures that a payment request is only considered valid when a majority of nodes agree, ensuring the consistency and accuracy of payment decisions. This threshold effectively prevents attacks from a small number of malicious nodes. Even if some nodes attempt to tamper with or deny a payment request, as long as the number of nodes that approve it exceeds the threshold, the payment request will still be considered valid, ensuring the security and stability of the payment system.

[0042] AI glasses consist of hardware, software, network, and application layers. Specifically, the hardware layer includes a camera, microphone, processor, display, battery, wireless communication modules (such as Wi-Fi, Bluetooth, and NFC), and a security chip. The camera captures iris images and gestures, the microphone captures voice, the processor runs various algorithms and applications, the display displays payment information and prompts, the battery provides power, the wireless communication module communicates with the payment server, and the security chip stores the user's private key and other sensitive information. The software layer includes an operating system, payment application, and security module. The operating system, based on a smart wearable device operating system such as Wear OS, provides the basic operating environment for AI glasses payments. The payment application, developed specifically for AI glasses, integrates functional modules such as multimodal biometric recognition, dynamic password authentication, end-to-end encryption, and blockchain technology to implement the core payment logic. The security module manages user private keys, certificates, and other security information, as well as performs encryption and decryption operations. At the network layer, communication between the AI glasses and the payment server is achieved through wireless communication networks (such as Wi-Fi, 4G / 5G, Bluetooth, etc.), ensuring the reliable transmission of payment information. At the application layer, payment-related application services are provided to users, such as payment record query, transaction details viewing, payment settings, etc., and it also connects with third-party payment platforms to realize payment settlement functions.

[0043] The embodiments of the present application have also been tested and verified. Functional testing: Comprehensive testing of functions such as multimodal biometric recognition, dynamic password verification, end-to-end encryption and blockchain technology to ensure the correctness and stability of each function. Security testing: Conduct security vulnerability scanning, penetration testing, etc. to detect whether the system has security risks, such as data leakage, identity fraud, etc. Evaluate the security of the encryption algorithm to ensure the confidentiality and integrity of payment information. Performance testing: Test performance indicators such as the response time and concurrent processing capabilities of AI glasses payment to ensure the stability and reliability of the payment system under high concurrency. User experience testing: Invite users to conduct actual usage tests, collect user feedback, and evaluate and improve aspects such as identity authentication processes and payment operation convenience.

[0044] The AI Glasses payment app will be pushed to users through the app store, allowing them to download and install it wirelessly. Simultaneously, the payment server will be deployed and configured to ensure stability and security. A comprehensive system maintenance and update mechanism will be established to regularly scan and fix security vulnerabilities in the AI Glasses payment system, and to promptly update the payment app's functionality and performance. Furthermore, user feedback will be collected to optimize and improve the system.

[0045] This embodiment also discloses a system for secure payment using AI glasses based on multiple encryption. Figure 2 This is a module diagram of the system for secure payment using AI glasses based on multiple encryption disclosed in the embodiment of this application, such as Figure 2 As shown, the system includes a collection module 201, a request module 202, a sending module 203 and a receiving module 204, wherein: A collection module 201 is configured to collect multimodal biometric information of the user in response to a user's operation of initiating a payment function, and determine whether the multimodal biometric information is consistent with preset biometric information, the multimodal biometric information including an iris image, a voice command, and a gesture image; The request module 202 is configured to send a dynamic password request to the bound target device when the multimodal biometric information is consistent with the preset biometric information, and receive and display the dynamic password returned by the target device on the display interface; The sending module 203 is configured to encrypt the payment information using the private key to obtain encrypted information when the user enters the dynamic password on the payment interface, and send the encrypted information to the payment server for decryption by the payment server and verification of the decrypted data through the blockchain network. When the verification result is passed, the funds are transferred according to the payment information; The receiving module 204 is configured to receive the fund transfer information returned by the payment server.

[0046] Optionally, the acquisition module 201 is configured to: Capturing an iris image through a camera, processing the iris image based on a lightweight convolutional neural network to obtain an encrypted feature vector, and calculating the cosine similarity between the encrypted feature vector and a preset feature vector; Collecting a voice signal through a microphone, processing the voice signal based on a Mel-frequency cepstral coefficient algorithm to obtain multidimensional feature parameters, and calculating a voiceprint feature dynamic time warping score based on the multidimensional feature parameters and a preset voiceprint template; Capturing gesture trajectories through a dynamic visual sensor, processing the gesture trajectories based on a long short-term memory network to obtain a spatiotemporal coding sequence, and calculating the gesture trajectory edit distance based on the spatiotemporal coding sequence and a preset gesture template; It is determined whether the cosine similarity, the voiceprint feature dynamic time warping score, and the gesture trajectory edit distance all meet preset conditions.

[0047] Optionally, the acquisition module 201 is configured to: Performing a fast Fourier transform and a discrete cosine transform on each frame of the speech signal to obtain a basic Mel-frequency cepstral coefficient feature, calculating a first-order difference and a second-order difference of the basic Mel-frequency cepstral coefficient feature, and expanding the basic Mel-frequency cepstral coefficient feature into a multidimensional feature parameter according to the first-order difference and the second-order difference; An MFCC feature sequence is constructed based on the multidimensional feature parameters corresponding to all frame signals of the speech signal, and the Mel-frequency cepstral coefficient feature sequence is compared with a preset voiceprint template to construct a comparison matrix. The optimal path from the upper left corner to the lower right corner in the comparison matrix is determined through dynamic programming. The distances of all matching frames are accumulated according to the optimal path to obtain a total cumulative distance. The total cumulative distance is divided by the number of all matching frames to obtain the dynamic time warping score of the voiceprint feature.

[0048] Optionally, the acquisition module 201 is configured to: Processing the spatial coordinates of the target finger in each frame image of the gesture trajectory to obtain a spatiotemporal coding vector, and constructing a spatiotemporal coding sequence based on all the spatiotemporal coding vectors of the gesture trajectory; A gesture feature dynamic time warping score is calculated based on the spatiotemporal coding sequence and the preset gesture template, and a gesture trajectory edit distance is determined based on the gesture feature dynamic time warping score.

[0049] Optionally, the request module 202 is configured to: Sending a dynamic password request to the bound target device through an elliptic curve encryption channel based on the device fingerprint, wherein the device fingerprint includes the device identification of the AI glasses; Receive the return data of the target device through the main channel, decrypt the return data to obtain the password plaintext, and display the dynamic password in plaintext on the display interface; A voice verification code is obtained through a backup channel, where the voice verification code is consistent with the dynamic password.

[0050] Optionally, the request module 202 is configured to: Determine a base point of the elliptic curve, generate a first random number as a private key of the elliptic curve, and generate a public key based on the private key and the base point. Binary concatenation of the device identifier and the private key to obtain concatenated data, performing a hash operation on the concatenated data to generate a hash value, and encoding the hash value to obtain a device fingerprint; A password request is generated according to the device fingerprint, the transaction information, and the second random number, and the password request is encrypted using the public key of the target device to obtain a dynamic password request.

[0051] Optionally, the sending module 203 is configured to: The encrypted information is decrypted using a public key algorithm in asymmetric encryption technology to obtain the payment information, the payment information is formatted according to the protocol requirements of the blockchain network to obtain a payment request, and the payment request is broadcast to multiple nodes in the blockchain network. The number of nodes that recognize the transaction legitimacy of the payment request is determined, and a ratio of the number to the total number of nodes is calculated. When the ratio exceeds a threshold, the payment request is confirmed to be legal.

[0052] It should be noted that the above embodiments provide devices that implement their functions using only the division of the above functional modules as examples. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiments are based on the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0053] This embodiment also discloses an electronic device, referring to Figure 3 The electronic device may include: at least one processor 301 , at least one communication bus 302 , a user interface 303 , a network interface 304 , and at least one memory 305 .

[0054] The communication bus 302 is used to implement the connection and communication between these components.

[0055] The user interface 303 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 303 may also include a standard wired interface and a wireless interface.

[0056] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0057] The processor 301 may include one or more processing cores. Using various interfaces and circuits, the processor 301 connects to various components within the server. It executes instructions, programs, code sets, or instruction sets stored in the memory 305, as well as accesses data stored in the memory 305, to perform various server functions and process data. Optionally, the processor 301 may be implemented using at least one of the following hardware forms: a digital signal processing (DSP), a field-programmable gate array (FPGA), or a programmable logic array (PLA). The processor 301 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing content displayed on the display screen; and the modem handles wireless communications. It is understood that the modem may not be integrated into the processor 301 but implemented as a separate chip.

[0058] Among them, the memory 305 may include a random access memory (RAM) or a read-only memory (Read-Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may also be optionally at least one storage device located away from the aforementioned processor 301. As Figure 3 As shown, the memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface module, and an application program for a method for secure payment of AI glasses based on multiple encryptions.

[0059] exist Figure 3In the electronic device shown, the user interface 303 is mainly used to provide an input interface for the user and obtain data input by the user; and the processor 301 can be used to call the application of the method of secure payment of AI glasses based on multiple encryption stored in the memory 305. When executed by one or more processors 301, the electronic device executes one or more methods as in the above embodiments.

[0060] It should be noted that for the aforementioned method embodiments, for simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for this application.

[0061] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0062] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are merely schematic, such as the division of units, which is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interface, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0063] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0064] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0065] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of this application, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory 305 and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of this application. The aforementioned memory 305 includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a magnetic disk, or an optical disk.

[0066] The above is only an exemplary embodiment of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification, those skilled in the art will easily think of other embodiments of the present disclosure. This application is intended to cover any variations, uses or adaptive changes of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary technical means in the technical field that are not recorded in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A method for secure payment using AI glasses based on multiple encryption, characterized in that: Applied to AI glasses, the method includes: In response to a user's operation of activating a payment function, collecting multimodal biometric information of the user and determining whether the multimodal biometric information is consistent with preset biometric information, the multimodal biometric information including an iris image, a voice command, and a gesture image; When the multimodal biometric information is consistent with the preset biometric information, a dynamic password request is sent to the bound target device, and the dynamic password returned by the target device is received and displayed on the display interface; When the user enters the dynamic password on the payment interface, the payment information is encrypted using the private key to obtain encrypted information, and the encrypted information is sent to the payment server for decryption by the payment server and verification of the decrypted data through the blockchain network. When the verification result is passed, the funds are transferred according to the payment information; Receive the funds transfer information returned by the payment server.

2. The method for secure payment using AI glasses based on multiple encryption according to claim 1, characterized in that: The collecting of the multimodal biometric information of the user and determining whether the multimodal biometric information is consistent with the preset biometric information includes: Capturing an iris image through a camera, processing the iris image based on a lightweight convolutional neural network to obtain an encrypted feature vector, and calculating the cosine similarity between the encrypted feature vector and a preset feature vector; Collecting a voice signal through a microphone, processing the voice signal based on a Mel-frequency cepstral coefficient algorithm to obtain multidimensional feature parameters, and calculating a voiceprint feature dynamic time warping score based on the multidimensional feature parameters and a preset voiceprint template; Capturing gesture trajectories through a dynamic visual sensor, processing the gesture trajectories based on a long short-term memory network to obtain a spatiotemporal coding sequence, and calculating the gesture trajectory edit distance based on the spatiotemporal coding sequence and a preset gesture template; It is determined whether the cosine similarity, the voiceprint feature dynamic time warping score, and the gesture trajectory edit distance all meet preset conditions.

3. The method for secure payment using AI glasses based on multiple encryption according to claim 1, characterized in that: The processing of the speech signal based on the Mel-frequency cepstral coefficient algorithm to obtain multidimensional feature parameters, and calculating the voiceprint feature dynamic time warping score according to the multidimensional feature parameters and a preset voiceprint template include: Performing a fast Fourier transform and a discrete cosine transform on each frame of the speech signal to obtain a basic Mel-frequency cepstral coefficient feature, calculating a first-order difference and a second-order difference of the basic Mel-frequency cepstral coefficient feature, and expanding the basic Mel-frequency cepstral coefficient feature into a multidimensional feature parameter according to the first-order difference and the second-order difference; An MFCC feature sequence is constructed based on the multidimensional feature parameters corresponding to all frame signals of the speech signal, and the Mel-frequency cepstral coefficient feature sequence is compared with a preset voiceprint template to construct a comparison matrix. The optimal path from the upper left corner to the lower right corner in the comparison matrix is determined through dynamic programming. The distances of all matching frames are accumulated according to the optimal path to obtain a total cumulative distance. The total cumulative distance is divided by the number of all matching frames to obtain the dynamic time warping score of the voiceprint feature.

4. The method for secure payment using AI glasses based on multiple encryption according to claim 1, characterized in that: The processing of the gesture trajectory based on the long short-term memory network to obtain a spatiotemporal coding sequence, and calculating the gesture trajectory edit distance according to the spatiotemporal coding sequence and a preset gesture template include: Processing the spatial coordinates of the target finger in each frame image of the gesture trajectory to obtain a spatiotemporal coding vector, and constructing a spatiotemporal coding sequence based on all the spatiotemporal coding vectors of the gesture trajectory; A gesture feature dynamic time warping score is calculated based on the spatiotemporal coding sequence and the preset gesture template, and a gesture trajectory edit distance is determined based on the gesture feature dynamic time warping score.

5. The method for secure payment using AI glasses based on multiple encryption according to claim 1, characterized in that: The sending of a dynamic password request to a bound target device, and receiving and displaying the dynamic password returned by the target device on a display interface includes: Sending a dynamic password request to the bound target device through an elliptic curve encryption channel based on the device fingerprint, wherein the device fingerprint includes the device identification of the AI glasses; Receive the return data of the target device through the main channel, decrypt the return data to obtain the password plaintext, and display the dynamic password in plaintext on the display interface; A voice verification code is obtained through a backup channel, where the voice verification code is consistent with the dynamic password.

6. The method for secure payment using AI glasses based on multiple encryption according to claim 5, characterized in that: The sending of a dynamic password request to a bound target device through an elliptic curve encryption channel according to the device fingerprint includes: Determine a base point of the elliptic curve, generate a first random number as a private key of the elliptic curve, and generate a public key based on the private key and the base point. Binary concatenation of the device identifier and the private key to obtain concatenated data, performing a hash operation on the concatenated data to generate a hash value, and encoding the hash value to obtain a device fingerprint; A password request is generated according to the device fingerprint, the transaction information, and the second random number, and the password request is encrypted using the public key of the target device to obtain a dynamic password request.

7. The method for secure payment using AI glasses based on multiple encryption according to claim 1, characterized in that: The sending of the encrypted information to the payment server for decryption by the payment server and verification of the decrypted data through the blockchain network includes: The encrypted information is decrypted using a public key algorithm in asymmetric encryption technology to obtain the payment information, the payment information is formatted according to the protocol requirements of the blockchain network to obtain a payment request, and the payment request is broadcast to multiple nodes in the blockchain network. The number of nodes that recognize the transaction legitimacy of the payment request is determined, and a ratio of the number to the total number of nodes is calculated. When the ratio exceeds a threshold, the payment request is confirmed to be legal.

8. A system for secure payment using AI glasses based on multiple encryption, characterized in that: It includes acquisition module, request module, sending module and receiving module, among which: a collection module configured to collect multimodal biometric information of the user in response to an operation by the user to activate a payment function, and determine whether the multimodal biometric information is consistent with preset biometric information, the multimodal biometric information including an iris image, a voice command, and a gesture image; a request module configured to send a dynamic password request to a bound target device when the multimodal biometric information is consistent with the preset biometric information, and receive and display the dynamic password returned by the target device on a display interface; a sending module configured to, when the user enters the dynamic password on the payment interface, encrypt the payment information using a private key to obtain encrypted information, and send the encrypted information to the payment server for decryption by the payment server and verification of the decrypted data through the blockchain network; when the verification result is passed, execute the fund transfer according to the payment information; The receiving module is configured to receive the funds transfer information returned by the payment server.

9. An electronic device, characterized in that: The electronic device comprises a processor, a memory, a user interface and a network interface, wherein the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 7 is executed.

Images