Identity recognition method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides an identity recognition method and device, electronic equipment and a storage medium, and relates to the technical field of computers. Comprising the steps of obtaining a target biological feature of a target mode of a to-be-identified user; mapping the target biological characteristics to a preset quantum multi-dimensional space to obtain quantum biological characteristics; and inputting the quantum biological characteristics into a pre-trained quantum identity recognition model to obtain an identity recognition result of the to-be-recognized user. Identity recognition is performed on the quantum biological characteristics based on the quantum identity recognition model obtained by training the quantum biological characteristics fusing multiple modalities, and on the basis of improving attack resistance of a system, dependence on a single modal can be reduced, so that the robustness of identity recognition is improved, and the robustness of identity recognition is improved. And meanwhile, the biological characteristics do not need to be compared with a large number of biological characteristics one by one, so that the identity recognition efficiency can be greatly improved.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of computer, and particularly relates to a method and device for identity recognition, electronic equipment and storage medium. BACKGROUND

[0002] Identity recognition using biological features has gradually become popular, and the biological features may include fingerprint features, facial image features, iris features, etc.

[0003] In the related art, biological feature data is usually directly stored in a storage device (such as a database) in the form of original data or encoded data of the original data, for example, facial image features can be stored in the form of a pixel matrix, and fingerprint features are stored in a specific image format.

[0004] When identity recognition is performed, the biological features of a user to be recognized are extracted and matched with the biological features in the data stored in the storage device, so as to obtain an identity recognition result. Taking fingerprint recognition as an example, the related art extracts the detail feature points (such as end points and bifurcation points) of the fingerprint of the user to be recognized, and then one-by-one compares the fingerprint features in the stored data, to determine whether they are the same fingerprint by calculating a similarity score.

[0005] The identity recognition technical solution in the related art has the following technical problems: 1. Poor robustness, when a biological feature of a certain modality is disturbed or damaged, it is often impossible to perform recognition, for example, if a face image is partially blocked, biological feature recognition cannot be performed.

[0006] 2. Low identity recognition efficiency The identity recognition in the related art needs to one-by-one compare a large amount of original data in the storage device (database), which consumes a long time, and the comparison efficiency will be significantly reduced when the amount of original data is large. Especially when high-dimensional data is processed, the calculation time will be greatly increased.

[0007] 3. High risk of data leakage Traditional storage devices (such as databases) are easy targets for network attacks, and hackers can use various means, such as Structured Query Language (SQL) injection, malicious software, password cracking, etc., to obtain access to the storage device (such as a database), and thus steal the stored biological feature data. SUMMARY

[0008] The purpose of the embodiments of the present disclosure is to provide a method and device for identity recognition, electronic equipment and storage medium.

[0009] To solve the above technical problems, the embodiments of the present disclosure are implemented through the following aspects.

[0010] According to a first aspect of the embodiments of the present disclosure, a method for identity recognition is provided, which comprises: obtaining a target biological feature of a target modality of a user to be identified; mapping the target biological feature to a preset quantum multidimensional space to obtain a quantum biological feature; inputting the quantum biological feature into a pre-trained quantum identity recognition model to obtain an identity recognition result of the user to be identified; wherein the quantum identity recognition model is trained by a fused quantum biological feature obtained by mapping biological features of multiple modalities corresponding to training users to the quantum multidimensional space, and the target modality is at least one modality in the multiple modalities.

[0011] According to a second aspect of the embodiments of the present disclosure, an apparatus for identity recognition is provided, which comprises: an obtaining module configured to obtain a target biological feature of a target modality of a user to be identified; a mapping module configured to map the target biological feature to a preset quantum multidimensional space to obtain a quantum biological feature; and an identifying module configured to input the quantum biological feature into a pre-trained quantum identity recognition model to obtain an identity recognition result of the user to be identified; wherein the quantum identity recognition model is trained by a fused quantum biological feature obtained by mapping biological features of multiple modalities corresponding to training users to the quantum multidimensional space, and the target modality is at least one modality in the multiple modalities.

[0012] According to a third aspect of the embodiments of the present disclosure, an electronic device is provided, which comprises: a processor; a memory for storing instructions executable by the processor; wherein the processor is configured to execute the steps of the method for identity recognition according to the first aspect.

[0013] According to a fourth aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, which stores one or more programs, and the one or more programs, when executed by an electronic device comprising multiple application programs, cause the electronic device to execute the steps of the method for identity recognition according to the first aspect.

[0014] One of the above technical solutions has the following advantages or beneficial effects: based on the quantum identity recognition model trained by the fused quantum biological features of multiple modalities, the identity recognition of the quantum biological features can improve the system's resistance to attacks, reduce the dependence on a single modality, improve the robustness of identity recognition, and does not need to compare the biological features with a large number of biological features one by one, which can greatly improve the efficiency of identity recognition.

[0015] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure.

[0016] Other features and advantages of the present disclosure will be described in detail in the following detailed implementation part. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments described in the present disclosure, and other drawings can be obtained by those skilled in the art without creative labor.

[0018] Figure 1 A flowchart of a method for identity recognition provided by an embodiment of the present disclosure is shown; Figure 2 Another flowchart of a method for identity recognition provided by an embodiment of the present disclosure is shown; Figure 3 Another flowchart of a method for identity recognition provided by an embodiment of the present disclosure is shown; Figure 4 Another flowchart of a method for identity recognition provided by an embodiment of the present disclosure is shown; Figure 5 A block diagram of an apparatus for identity recognition provided by an embodiment of the present disclosure is shown; Figure 6 Another block diagram of an apparatus for identity recognition provided by an embodiment of the present disclosure is shown; Figure 7 A hardware structure diagram of an electronic device for executing the method for identity recognition provided by an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0019] In order to make the person skilled in the art better understand the technical solutions in the present disclosure, the technical solutions in the embodiments of the present disclosure will be described clearly and completely in the following with reference to the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only some embodiments of the present disclosure, not all embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present disclosure.

[0020] The technical solutions of the present application will be described below in conjunction with specific embodiments.

[0021] Figure 1 A flowchart of a method for identity recognition provided by an embodiment of the present disclosure is shown, as shown in Figure 1 The method can include the following steps: In step S101, a target biological feature of a target modality of a user to be identified is acquired.

[0022] The modality in the present application can be understood as a category of biological features of the user to be identified acquired by a corresponding acquisition device. For example, the fingerprint biological feature acquired by a fingerprint acquisition device is the biological feature of the fingerprint modality, and the facial image biological feature acquired by an image acquisition device is the biological feature of the image modality.

[0023] The target modality described above can be one modality, for example, the fingerprint modality, or multiple modalities. For example, the target biological feature can include the biological feature of the fingerprint modality and the biological feature of the image modality of the user to be identified, and the present application does not limit this.

[0024] In step S102, the target biological feature is mapped to a preset quantum multidimensional space to obtain a quantum biological feature.

[0025] In step S103, the quantum biological feature is input into a pre-trained quantum identity recognition model to obtain an identity recognition result of the user to be identified.

[0026] The quantum identity recognition model is trained by a fusion quantum biological feature obtained by mapping multiple modalities of biological features corresponding to a training user to a quantum multidimensional space, and the target modality is at least one modality in the multiple modalities.

[0027] For example, when the quantum identity recognition model is trained, the multiple modalities corresponding to the training user can be a complete set of different modalities. For example, the multiple modalities of biological features corresponding to the training user can include fingerprint features, image features, and iris features, and the fusion quantum biological feature is obtained by mapping the biological features of the multiple modalities to a quantum multidimensional space.

[0028] In some embodiments, the biological features of the multiple modalities include at least one biological feature corresponding to each modality, each biological feature corresponds to a quantum bit, and the quantum multidimensional space is a Hilbert space formed by the combination of quantum bits corresponding to the biological features of the multiple modalities.

[0029] Each quantum bit is in a superposition state of and The quantum bit can be represented by the following formula one.

[0030] (Formula one) Wherein, and are the basis vectors of quantum computing, and are the quantum bits collapse to and the probability that is a modulo operation.

[0031] The quantum multi-dimensional space is formed by corresponding quantum bit combinations of n-dimensional Hilbert space, for example, the fingerprint feature corresponding to each training user is an m1-dimensional vector, the image feature is an m2-dimensional vector, and the iris feature is an m3-dimensional vector, then the fingerprint feature corresponds to m1 quantum bits, the image feature corresponds to m2 quantum bits, and the iris feature corresponds to m3 quantum bits, and The n-dimensional Hilbert space is obtained by tensor product (i.e., Kronecker product) of m1+m2+m3 quantum bits.

[0032] After inputting the quantum biological features of the user to be identified into the quantum identity recognition model, since the quantum identity recognition model is trained according to the fused quantum biological features, which fuse quantum biological features of multiple modalities, even if the target biological features of the user to be identified only include biological features of any one modality (such as fingerprint features, image features), the identity of the user to be identified can also be accurately identified.

[0033] By using the above technical solution, the quantum identity recognition model trained based on the fused quantum biological features of multiple modalities is used for identity recognition of quantum biological features, which can reduce the dependence on a single modality on the basis of improving the resistance to attacks, thereby improving the robustness of identity recognition, and without the need for one-by-one comparison between biological features and a large number of biological features, the identity recognition efficiency can be greatly improved.

[0034] Figure 2 Another flowchart of the method for identity recognition provided by the embodiments of the present disclosure is shown in FIG. 2. Figure 2 As shown in FIG. 2, step S102 can specifically include the following steps.

[0035] In step S1021, the target biological features are respectively mapped to corresponding quantum bits in the quantum multi-dimensional space.

[0036] In some embodiments, the mapping function shown in Formula Two below can be used to map the feature vectors of different modalities to the coefficients of the quantum multi-dimensional space .

[0037] (Formula Two) Taking the image feature as the target biological feature, the image feature vector includes two feature dimensions, which can be represented as The mapping function shown in Formula Three below can be used to respectively map the target biological features to corresponding quantum bits in the quantum multi-dimensional space, to obtain the mapped quantum state.

[0038] (Formula Three) in, Represents the tensor product.

[0039] After mapping the image feature vector to the corresponding qubit in quantum multidimensional space, when the target biometric includes multiple modalities (e.g., fingerprint modality), the quantum state after fingerprint modality mapping can be obtained by referring to Equations 2 and 3 above. Then through calculation and The tensor product maps the multiple target biofeatures to their corresponding qubits in the quantum multidimensional space, resulting in the mapped quantum state.

[0040] In step S1022, quantum operations are performed on the quantum state mapped to the quantum multidimensional space to quantum fuse the target biofeatures to obtain quantum biofeatures.

[0041] In some embodiments, quantum operations include Hadamard gate quantum operations and controlled NOT gate quantum operations.

[0042] Among them, the Hadamard gate can transform a qubit from its ground state (e.g., ... and A qubit can be converted from a superposition state to a ground state. The definition of an n-order Hadamard gate is determined by the following formula four.

[0043] (Formula 4) Controlled-NOT gates, also known as CNOT gates, can create entangled qubits. For example, for two corresponding quantum states... and Quantum bits, controlled NOT gates can for In the case of, then remain unchanged. Change to X ,otherwise and All remain unchanged. That is... .

[0044] The target biometrics include two modalities: fingerprint features and image features (understandably, the number of modalities included in the target biometrics can be less than the number of modalities included in the biometrics corresponding to the training user). The fingerprint feature is a three-dimensional feature vector, and the image feature is a four-dimensional feature vector. Therefore, the feature vector dimension of the user to be identified is 3 + 4 = 7 dimensions. In step S1021, the 7-dimensional feature vector can be mapped to the corresponding 7 qubits. If the quantum identity recognition model is obtained by training a fused quantum biometric feature obtained by mapping the biometrics of the two modalities corresponding to the training user to a quantum multidimensional space, then this quantum multidimensional space can be... The Hilbert space of dimension 3. Through the above steps S1021 and S1022, the target biometric features are mapped to... Quantum biological characteristics are obtained in a dimensional Hilbert space. In some possible implementations, one could... Basis vectors of a dimensional Hilbert space (e.g., one of the basis vectors could be...) The 128 coefficients of ) are used as quantum biological characteristics.

[0045] By adopting the above technical solution, the biometric features of the user to be identified can be mapped to a cross-modal quantum multidimensional space, which makes it easy to input the obtained fused quantum biometric features of the user to be identified into a pre-trained quantum identity recognition model to obtain the identity recognition result of the user. This can reduce the dependence on a single modality, thereby improving the robustness of identity recognition. At the same time, it does not require comparing biometric features with a large number of biometric features one by one, which can significantly improve the efficiency of identity recognition.

[0046] In some embodiments, the target biometrics of the target user can be obtained by performing feature dimensionality reduction on the original biometrics of the collected target modality.

[0047] Figure 3 This illustration shows yet another flowchart of the identity recognition method provided in an embodiment of the present disclosure, such as... Figure 3 As shown, step S101 may include the following steps.

[0048] In step S1011, the covariance matrix of the original biometric features is constructed.

[0049] For example, suppose a target modality (e.g., fingerprint modality) has m original biometric features, each of which is an n-dimensional feature vector, forming an m-row n-column original biometric feature matrix X. Then the covariance matrix C is an n-row n-column square matrix, where the elements of the covariance matrix C can be represented by the following formula five.

[0050] (Formula 5) in, the mean value of the pth feature, is the element value of the ith row and the pth column of the original biological feature matrix X.

[0051] In step S1012, eigenvalue decomposition is performed on the covariance matrix to obtain eigenvalues and eigenvectors of the original biological features.

[0052] For example, the eigenvalue decomposition can be performed on the covariance matrix according to Formula Six as follows, (Formula Five) wherein, is a diagonal matrix whose diagonal elements are the eigenvalues, and V is an eigenmatrix composed of the eigenvectors, is the transpose of the eigenmatrix.

[0053] In step S1013, the original biological features are reduced in dimension according to the target eigenvector to obtain target biological features.

[0054] wherein the target eigenvector is an eigenvector in which the corresponding eigenvalue is greater than or equal to a preset threshold. The eigenvalue reflects the importance of the feature represented by the corresponding eigenvector, and a larger eigenvalue contains more important information. For example, in fingerprint feature extraction, if the eigenvalue corresponding to an eigenvector is large, it indicates that the fingerprint texture feature represented by the eigenvector is more representative in the entire set of fingerprint features.

[0055] In some embodiments, the original biological features are obtained by digitizing the corresponding image data (such as face images, fingerprint images, and iris images) collected by a collection device (such as a sensor or a scanner) and extracting corresponding image features.

[0056] In some embodiments, in order to further improve the quality of the original biological features, image denoising processing can be performed on the image using image filtering techniques such as Gaussian filtering before the image features are extracted. For example, by performing weighted summation on each pixel point and its neighborhood pixel points in the image, Gaussian noise in the image can be effectively removed.

[0057] In some embodiments, the collected biological feature data can also be normalized (for example, pixel values 0-255 are normalized to the range 0-1) to facilitate the fusion of multi-modal biological features.

[0058] By using the above technical solutions, the original biological features are reduced in dimension according to the eigenvectors in which the corresponding eigenvalues are greater than or equal to a preset threshold, which can reduce the dimension on the basis of retaining important biological features, further reduce the computational load, and improve the efficiency of identity recognition.

[0059] It can be understood that the identity recognition result can have different application modes, for example, it can be a classification task of determining whether the to-be-identified user has the permission to enter a certain area, and it can also be an identity recognition task of outputting the identity information corresponding to the to-be-identified user. Correspondingly, the quantum identity recognition model can also be a machine learning model of different types. For example, it can be a quantum multidimensional space forest model, which can include a plurality of quantum decision tree models for processing classification tasks, and each quantum decision tree model can output whether the to-be-identified user has the permission to enter a certain area. Of course, the quantum identity recognition model can also be a neural network model for outputting the identity information corresponding to the to-be-identified user. According to the different machine learning models, the labels corresponding to the fusion quantum biological features obtained after the biological features of the training user corresponding to a plurality of modalities are mapped to the quantum multidimensional space are also different. For example, if it is a quantum multidimensional space forest model for processing classification tasks, the label can be "yes" or "no".

[0060] Figure 4 Another flowchart of the method of identity recognition provided by the embodiments of the present disclosure is shown in FIG. 10. Figure 4 As shown in FIG. 10, taking the quantum multidimensional space forest model as an example, the quantum multidimensional space forest model is trained by the following method: In step S201, samples are randomly drawn from the quantum multidimensional space dataset composed of the fusion quantum biological features of the training user and the corresponding label data.

[0061] When the quantum multidimensional space forest model is trained, the plurality of modalities corresponding to the training user can be the full set of different modalities.

[0062] In some embodiments, the biological features of each modality used to train the quantum identity recognition model can include a plurality of groups of biological features, and each group of biological features includes at least one biological feature for biological recognition. For example, the biological features of the fingerprint modality can include a plurality of groups of biological features obtained by performing rotation, flipping and other transformation operations on the fingerprint images of the same finger and then extracting features. The biological features of the image modality can include a plurality of groups of biological features obtained by performing rotation, flipping and other transformation operations on the face images and then extracting features. In this way, even if the face image or the fingerprint image of the to-be-identified user is damaged (for example, partially blocked), biological feature recognition can still be accurately performed.

[0063] In some possible implementations, the quantum multidimensional space dataset composed of the fusion quantum biological features and the corresponding label data can be obtained by the following steps.

[0064] In step 10, the biological features of the plurality of modalities corresponding to each training user are obtained.

[0065] In some embodiments, the original biological features of the collected training users can be subjected to feature dimension reduction by adopting steps similar to steps S1011-S1013, and the biological features of the training users in multiple modalities after dimension reduction can be obtained, so as to reduce the amount of calculation and improve the efficiency of model training and subsequent identity recognition.

[0066] In step 11, the biological features in multiple modalities can be mapped to the preset quantum multidimensional space to obtain the fused quantum biological features in a similar manner to step S102.

[0067] In step 12, the obtained fused quantum biological features are corresponded with the label data, so as to obtain a quantum multidimensional space data set composed of the fused quantum biological features and the corresponding label data.

[0068] In step S202, the fused quantum biological features of the samples are divided according to the quantum information entropy reduction criterion, so as to obtain multiple quantum decision tree models.

[0069] At each node of the quantum decision tree, the division is performed according to the quantum features. The quantum information entropy is calculated according to Formula Six below to determine the optimal division feature.

[0070] (Formula Six) Wherein, S is the quantum information entropy, is the probability of the quantum state being in the i-th possible state, the optimal division feature is selected by comparing the information entropy reduction amount when the decision tree is divided according to different quantum features to generate a branch of the decision tree, and the above steps are repeated, when the information entropy is reduced to a certain extent or reaches a preset tree depth, the division is stopped, and a quantum decision tree is constructed, and in this way, multiple quantum decision tree models can be constructed.

[0071] In some possible implementation manners, the data set can be divided into a training set and a validation set, for example, the data set can be divided according to a ratio of 8:2, and the quantum decision tree model constructed is pre-pruned and / or post-pruned according to the data of the validation set, so as to avoid overfitting of the quantum decision tree.

[0072] In the training process, the model parameters are adjusted by minimizing the loss function (such as the cross loss function), and for each quantum decision tree, the quantum feature division threshold and node structure inside it can be adjusted according to the training data to improve the accuracy of the model.

[0073] In step S203, the quantum decision tree models are combined to obtain a quantum multidimensional space forest model.

[0074] In some embodiments, the plurality of quantum decision tree models generated in step S202 can be combined to obtain a quantum multidimensional space forest model. In the prediction stage, each quantum decision tree independently predicts the input quantum biological features, and then integrates the prediction results. By combining the plurality of quantum decision tree models, the prediction results of the plurality of quantum decision trees are integrated to make a fusion decision (for example, a voting method can be used to determine the final identity recognition result. If the results of the plurality of quantum decision trees are the same, the final recognition result is determined), which can obtain significantly superior generalization performance than a single quantum decision tree model.

[0075] In some possible implementations, a hyperparameter adjustment method can be used to set different value ranges for important hyperparameters of the quantum multidimensional space forest model, such as the number of trees and the maximum depth of each tree, and find the best combination of hyperparameters on the validation set by exhaustively combining these values. At the same time, model fusion techniques in ensemble learning, such as fusing multiple quantum multidimensional space forest models initialized differently or set with different hyperparameters, are used to further improve the performance and stability of the model.

[0076] It can be understood that the quantum identity recognition model can also be other types of machine learning models, which can be obtained according to the training method of the model in the related art based on the quantum multidimensional space data set composed of the fused quantum biological features of the training user and the corresponding label data. The specific training method is not described again.

[0077] By using the technical solutions described above, a quantum identity recognition model can be trained, so that the quantum biological features can be identified based on the quantum identity recognition model, the robustness of identity recognition is improved on the basis of improving the system's resistance to attacks, and the identity recognition efficiency can be greatly improved without comparing the biological features with a large number of biological features one by one.

[0078] Figure 5 A block diagram of an identity recognition apparatus provided by an embodiment of the present application is shown in FIG. 1. Figure 5 As shown in FIG. 1, the identity recognition apparatus 100 includes: An acquisition module 110 is configured to acquire target biological features of a target modality of a user to be identified. A mapping module 120 is configured to map the target biological features to a preset quantum multidimensional space to obtain quantum biological features. An identification module 130 is configured to input the quantum biological features into a pre-trained quantum identity recognition model to obtain an identity recognition result of the user to be identified. The quantum identity recognition model is trained based on fused quantum biological features obtained by mapping biological features of a plurality of modalities of a training user to a quantum multidimensional space, and the target modality is at least one of the plurality of modalities.

[0079] Optionally, the biometrics of multiple modalities include at least one biometric corresponding to each modality, each biometric corresponding to a qubit, and the quantum multidimensional space is a Hilbert space formed by the combination of qubits corresponding to the biometrics of multiple modalities.

[0080] Optionally, the mapping module 120 is further configured to: map the target biometric features to corresponding qubits in the quantum multidimensional space; Quantum operations are performed on quantum states mapped to quantum multidimensional space to quantumly fuse target biosignatures and obtain quantum biosignatures.

[0081] Optionally, quantum operations include Hadamard gate quantum operations and controlled NOT gate quantum operations.

[0082] Optionally, the acquisition module 110 is also used to perform feature dimensionality reduction on the original biometric features of the acquired target modality to obtain the target biometric features of the target modality of the user to be identified.

[0083] Optionally, the acquisition module 110 is also used for: Construct the covariance matrix of the original biological features; Eigenvalue decomposition of the covariance matrix yields the eigenvalues ​​and eigenvalue vectors of the original biological features; The target biological feature is obtained by dimensionality reduction of the original biological feature based on the target feature value vector. The target feature value vector is the feature value vector in which the corresponding feature value is greater than or equal to a preset threshold.

[0084] The device 100 provided in this application embodiment can execute the methods in the preceding method embodiments, realize the functions of the methods in the preceding method embodiments, and achieve the corresponding beneficial effects, which will not be repeated here.

[0085] Figure 6 This diagram illustrates a block diagram of another identity recognition device provided in an embodiment of this application, such as... Figure 6 As shown, the identity recognition device 100 also includes a training module 140, used for: Samples are randomly drawn with replacement from a quantum multidimensional space dataset consisting of fused quantum biometrics and corresponding label data of the training users; Multiple quantum decision tree models are obtained by fusing quantum biological characteristics to divide samples based on the reduction of quantum information entropy. The quantum decision tree model is combined to obtain the quantum multidimensional space forest model.

[0086] The device 100 provided in this application embodiment can execute the methods in the preceding method embodiments and realize the functions and beneficial effects of the methods in the preceding method embodiments, which will not be repeated here.

[0087] Figure 7 A hardware structure schematic diagram of an electronic device is shown, which is provided by the embodiments of the present disclosure, as shown in Figure 7 At the hardware level, the electronic device includes at least one processor, and optionally includes an internal bus, a network interface, and a memory. The memory can include an internal memory, such as a high-speed random access memory (RAM), and can also include a non-volatile memory, such as at least one disk memory. Of course, the electronic device can also include other hardware required by the business.

[0088] The processor, the network interface, and the memory can be connected to each other through the internal bus, which can be an industry standard architecture (ISA) bus, a peripheral component interconnect (PCI) bus, or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one bidirectional arrow is used in the figure, but it does not mean that there is only one bus or only one type of bus.

[0089] The memory stores programs. Specifically, the programs can include program codes, which include at least one computer operation instruction. The memory can include an internal memory and a non-volatile memory, and provide instructions and data to the processor.

[0090] The at least one processor reads the corresponding computer program from the non-volatile memory into the memory and then runs, and forms the device for positioning the target user at the logical level. The at least one processor executes the programs stored in the memory, and specifically executes the method disclosed in the embodiments of the first aspect and realizes the functions and beneficial effects of the methods described in the foregoing method embodiments, which will not be repeated here.

[0091] The method disclosed in the embodiment of the first aspect of the present disclosure can be applied to at least one processor or implemented by at least one processor. The processor can be an integrated circuit chip with a processing capability of signals. In the implementation process, each step of the above method can be completed by integrated logic circuits of hardware in the at least one processor or instructions in the form of software. The processor mentioned above can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. Each method, step and logic block diagram disclosed in the embodiment of the present disclosure can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in combination with the embodiment of the present disclosure can be directly embodied as a hardware code processor for execution, or a combination of hardware and software modules in the code processor for execution. The software module can be located in a random memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, a register or other mature storage medium in the art. The storage medium is located in the memory, and the processor reads the information in the memory and combines the hardware to complete the steps of the above method.

[0092] The electronic device can also perform the methods described in the foregoing method embodiments and achieve the functions and beneficial effects of the methods described in the foregoing method embodiments, which will not be repeated here.

[0093] Of course, in addition to the software implementation, the electronic device of the present disclosure does not exclude other implementation manners, such as logic devices or a combination of software and hardware, etc. That is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or a logic device.

[0094] The embodiment of the present disclosure further proposes a computer readable storage medium, the computer readable medium stores one or more programs, when the one or more programs are executed by at least one processor, the method disclosed in the embodiment of the first aspect is implemented and the functions and beneficial effects of the methods described in the foregoing method embodiments are achieved, which will not be repeated here.

[0095] The computer readable storage medium includes a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, etc.

[0096] Further, the embodiment of the present disclosure also provides a computer program product, which comprises a computer program stored on a non-transitory computer readable storage medium, and the computer program comprises program instructions, when the program instructions are executed by a computer, the following processes are implemented: the method disclosed in the embodiment of the first aspect and the functions and beneficial effects of the methods described in the foregoing method embodiments, which are not described here again.

[0097] In summary, the above only describes the preferred embodiments of the present disclosure, and does not limit the protection scope of the present disclosure. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

[0098] The system, module or unit illustrated in the above embodiments can be specifically implemented by a computer chip or entity, or by a product with certain functions. A typical implementation device is a computer. Specifically, the computer may, for example, be a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0099] The computer readable medium includes permanent and non-permanent, removable and non-removable media, which can be implemented by any method or technology to store information. The information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can store information accessible by a computing device. According to the definition herein, the computer readable medium does not include transitory computer readable media, such as modulated data signals and carriers.

[0100] It is also to be noted that the terms "comprising", "including", and any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises a... " does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0101] The various embodiments in the specification are described in progressive manner, and the same or similar parts among the various embodiments can be mutually referred to, and each embodiment focuses on the difference from other embodiments. In particular, for the system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the part of the method embodiments.

Claims

1. A method of identity recognition, characterized in that, The method comprises: acquiring a target biological feature of a target modality of a user to be identified; mapping the target biological feature to a preset quantum multidimensional space to obtain a quantum biological feature; inputting the quantum biological feature into a pre-trained quantum identity recognition model to obtain an identity recognition result of the user to be identified; wherein the quantum identity recognition model is trained by mapping biological features of multiple modalities corresponding to training users to the quantum multidimensional space to obtain fused quantum biological features, and the target modality is at least one of the multiple modalities.

2. The method of claim 1, wherein, The biological features of the multiple modalities include at least one biological feature corresponding to each modality, and each biological feature corresponds to a qubit. The quantum multidimensional space is a Hilbert space formed by combinations of qubits corresponding to the biological features of the multiple modalities.

3. The method of claim 2, wherein, The mapping of the target biological feature to the preset quantum multidimensional space to obtain the quantum biological feature comprises: mapping the target biological feature to corresponding qubits in the quantum multidimensional space respectively; performing quantum operations on quantum states mapped to the quantum multidimensional space to quantum fuse the target biological feature to obtain the quantum biological feature.

4. The method of claim 3, wherein, The quantum operations include Hadamard gate quantum operations and controlled non gate quantum operations.

5. The method of claim 1, wherein, The acquisition of the target biological feature of the target modality of the user to be identified comprises: performing feature dimension reduction on the original biological feature of the target modality collected to obtain the target biological feature of the target modality of the user to be identified.

6. The method of claim 5, wherein, The feature dimension reduction on the original biological feature of the target modality collected to obtain the target biological feature of the target modality of the user to be identified comprises: constructing a covariance matrix of the original biological feature; performing eigenvalue decomposition on the covariance matrix to obtain eigenvalues and eigenvectors of the original biological feature; performing dimension reduction on the original biological feature according to a target eigenvector to obtain the target biological feature, wherein the target eigenvector is an eigenvector in the eigenvector corresponding to an eigenvalue greater than or equal to a preset threshold.

7. The method according to any one of claims 1 to 6, characterized in that, The quantum identity recognition model is a quantum multidimensional space forest model, which is trained by the following method: randomly sampling from a quantum multidimensional space dataset composed of fused quantum biological features of training users and corresponding label data; dividing the fused quantum biological features of the samples according to quantum information entropy reduction as a standard to obtain multiple quantum decision tree models; combining the quantum decision tree models to obtain the quantum multidimensional space forest model.

8. An apparatus for identity recognition, the apparatus comprising: The device comprises: an acquisition module configured to acquire a target biological feature of a target modality of a user to be identified; a mapping module configured to map the target biological feature to a preset quantum multidimensional space to obtain a quantum biological feature; an identification module configured to input the quantum biological feature into a pre-trained quantum identity recognition model to obtain an identity recognition result of the user to be identified. The quantum identity recognition model is obtained by mapping a plurality of modal biological features corresponding to a training user to a fused quantum biological feature in a quantum multidimensional space.

9. An electronic device, comprising: Comprise: A memory, a processor, and a computer program stored on the memory and executable on the processor, the computer program, when executed by the processor, implements the identity recognition method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer program is stored on the computer readable storage medium, and the computer program, when executed by the processor, implements the identity recognition method of any one of claims 1 to 7.

11. A computer program product, characterised in that, The computer program product comprises a computer program stored on a non-transitory computer readable storage medium, and the computer program comprises program instructions, which, when executed by a computer, implement the identity recognition method of any one of claims 1 to 7.

Images