CN111950416B - Face recognition method and system based on block chain - Google Patents

Face recognition method and system based on block chain Download PDF

Info

Publication number
CN111950416B
CN111950416B CN202010762414.3A CN202010762414A CN111950416B CN 111950416 B CN111950416 B CN 111950416B CN 202010762414 A CN202010762414 A CN 202010762414A CN 111950416 B CN111950416 B CN 111950416B
Authority
CN
China
Prior art keywords
face recognition
model parameters
recognition model
blockchain
block chain
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202010762414.3A
Other languages
Chinese (zh)
Other versions
CN111950416A (en
Inventor
胡兴源
李艳
徐颖
周新衡
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Industrial and Commercial Bank of China Ltd ICBC
Original Assignee
Industrial and Commercial Bank of China Ltd ICBC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Industrial and Commercial Bank of China Ltd ICBC filed Critical Industrial and Commercial Bank of China Ltd ICBC
Priority to CN202010762414.3A priority Critical patent/CN111950416B/en
Publication of CN111950416A publication Critical patent/CN111950416A/en
Application granted granted Critical
Publication of CN111950416B publication Critical patent/CN111950416B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/60—Protecting data
    • G06F21/62—Protecting access to data via a platform, e.g. using keys or access control rules
    • G06F21/6218—Protecting access to data via a platform, e.g. using keys or access control rules to a system of files or objects, e.g. local or distributed file system or database
    • G06F21/6245—Protecting personal data, e.g. for financial or medical purposes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/60—Protecting data
    • G06F21/64—Protecting data integrity, e.g. using checksums, certificates or signatures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/50—Maintenance of biometric data or enrolment thereof
    • G06V40/53—Measures to keep reference information secret, e.g. cancellable biometrics
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Software Systems (AREA)
  • Bioethics (AREA)
  • Computer Security & Cryptography (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Hardware Design (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Computational Linguistics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Biophysics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Medical Informatics (AREA)
  • Databases & Information Systems (AREA)
  • Image Analysis (AREA)
  • Collating Specific Patterns (AREA)

Abstract

The invention provides a face recognition method and system based on a block chain, and belongs to the technical field of block chains. The face recognition method based on the block chain comprises the following steps: acquiring current facial image data of a user; inputting the current facial image data into an optimal face recognition model to obtain a current face recognition result; the step of creating the optimal face recognition model comprises the following steps: uploading the face recognition model parameters to a blockchain network so that the blockchain network generates global model parameters according to the face recognition model parameters; and creating an optimal face recognition model according to the global model parameters. The invention can improve the face recognition accuracy and protect the privacy data of the user.

Description

Face recognition method and system based on block chain
Technical Field
The invention relates to the technical field of blockchains, in particular to a face recognition method and system based on a blockchain.
Background
Deep learning belongs to a branch system of machine learning, and has made breakthrough progress in a plurality of application fields such as voice recognition, natural language processing, computer vision and the like in recent years. The application of the deep learning technology in the aspect of face recognition promotes the development of the field of computer vision, and brings great convenience to the life of people.
Face recognition is performed based on a deep learning technology, and generally, an example of a facial feature model is established by training facial image data of a person and extracting facial image features of an individual. The face recognition model constructed here can be used for individual identity authentication, such as user identity login confirmation by using face recognition by a login system. When using this face recognition model for authentication, for a given face picture, the model gives a predictive value to determine whether to log in for the user himself. When the model is used for face recognition, the prediction effect depends on the accuracy and generalization of the model instance. The deep learning algorithm needs a large amount of training data in the training process to obtain an ideal training effect, however, face recognition training data sets owned by single institutions are independent of each other and cannot be shared at present, so that a user needs to acquire face images again when using a model of each institution. In addition, the face image data often relates to user privacy, and each institution cannot directly share the face image data to other institutions without user authorization, so that the accuracy of face recognition cannot be improved by directly sharing the face image data set of the user.
The collaborative deep learning is a distributed deep learning method, and when face recognition is carried out by training face data by using the collaborative deep learning, participating institutions only need to train by locally using the face image data of the user owned by the institutions and upload model parameters to a centralized parameter server, and the parameter server carries out parameter aggregation. The method realizes the training data sharing effect between the institutions and simultaneously protects the privacy of facial image data of the local users of the institutions. However, research shows that unauthorized malicious participating institutions in collaborative deep learning can acquire privacy face image training data of other participating institutions through local training of an countermeasure generation network, so that privacy information of users is revealed.
Disclosure of Invention
The embodiment of the invention mainly aims to provide a face recognition method and a face recognition system based on a block chain, so that the accuracy of face recognition is improved and user privacy data is protected.
In order to achieve the above object, an embodiment of the present invention provides a face recognition method based on a blockchain, including:
acquiring current facial image data of a user;
inputting the current facial image data into an optimal face recognition model to obtain a current face recognition result; the step of creating the optimal face recognition model comprises the following steps: uploading the face recognition model parameters to a blockchain network so that the blockchain network generates global model parameters according to the face recognition model parameters; and creating an optimal face recognition model according to the global model parameters.
The embodiment of the invention also provides a face recognition system based on the block chain, which comprises the following steps:
an acquisition unit configured to acquire current face image data of a user;
the face recognition result unit is used for inputting the current face image data into the optimal face recognition model to obtain a current face recognition result; the step of creating the optimal face recognition model comprises the following steps: uploading the face recognition model parameters to a blockchain network so that the blockchain network generates global model parameters according to the face recognition model parameters; and creating an optimal face recognition model according to the global model parameters.
The embodiment of the invention also provides computer equipment, which comprises a memory, a processor and a computer program stored on the memory and running on the processor, wherein the processor realizes the steps of the face recognition method based on the blockchain when executing the computer program.
The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when being executed by a processor, realizes the steps of the face recognition method based on the blockchain.
The face recognition method and the face recognition system based on the blockchain can improve the face recognition accuracy and protect the privacy data of the user.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a blockchain-based face recognition method in an embodiment of the invention;
FIG. 2 is a schematic diagram of a block chain network and participating structures in an embodiment of the present invention;
FIG. 3 is a flow chart of creating an optimal face recognition model in an embodiment of the invention;
FIG. 4 is a schematic diagram of creating an optimal face recognition model in accordance with an embodiment of the present invention;
FIG. 5 is a block diagram of a blockchain-based face recognition system in an embodiment of the present invention;
fig. 6 is a block diagram of a computer device in an embodiment of the invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Those skilled in the art will appreciate that embodiments of the invention may be implemented as a system, apparatus, device, method, or computer program product. Accordingly, the present disclosure may be embodied in the following forms, namely: complete hardware, complete software (including firmware, resident software, micro-code, etc.), or a combination of hardware and software.
In view of the fact that the prior art cannot improve the face recognition accuracy and prevent the user privacy from being revealed, the embodiment of the invention provides a face recognition method based on a blockchain, so that the face recognition accuracy is improved and the user privacy data is protected. The present invention will be described in detail with reference to the accompanying drawings.
The blockchain is a chained storage structure which is connected in sequence, and ensures the consistency and the non-tamper property of stored data through a consensus mechanism. The intelligent contract is a section of automatically executed electronic contract code stored in the blockchain network, and a developer can write the contract code according to the requirements to complete corresponding functions. According to the invention, a centralized parameter server in the traditional collaborative deep learning is removed, a decentralised annular architecture is adopted among the participating mechanisms, and the communication efficiency among the participating mechanisms in the face image data training process is improved by using the mixed precision training. The invention uses the blockchain technology to protect the privacy of training data of the participating mechanism, and through a similarity detection algorithm, malicious participants are stopped to a certain extent, the privacy training data set of the participating mechanism is further protected, and the accuracy and generalization of a finally constructed face recognition model are improved; the deep learning process comprises N independent participating mechanisms, each participating mechanism is a block chain node at the same time, and the complete block information is stored locally. The participating institutions send and receive updated model parameter data through a blockchain network, and model parameter aggregation is performed by intelligent contracts running in the blockchain.
Fig. 1 is a flowchart of a blockchain-based face recognition method in an embodiment of the present invention. FIG. 2 is a schematic diagram of a block chain network and participating structures in an embodiment of the present invention. As shown in fig. 1-2, the face recognition method based on the blockchain includes:
s101: current facial image data of the user is acquired.
S102: and inputting the current facial image data into an optimal face recognition model to obtain a current face recognition result.
The step of creating the optimal face recognition model comprises the following steps: uploading the face recognition model parameters to a blockchain network so that the blockchain network generates global model parameters according to the face recognition model parameters; and creating an optimal face recognition model according to the global model parameters.
The current face recognition result is the probability of each user identity, and the user identity corresponding to the maximum probability is the user identity of the current face image data. There are a total of N blockchain nodes in the blockchain network, i.e., there are N participating mechanisms. As shown in fig. 2, n=5, then P is shared 1 -P 5 Five participating institutions are provided. The identity of the participating institution needs to be determined byThe authenticated user is a mechanism which is ensured to be legal and has the face image data of the user with a trusted source through an authentication system. For a particular model training task, all participating institutions negotiate a common training goal and use the same model training method. And constructing a blockchain alliance network between participating institutions, and constructing a server environment required by deep learning training, such as training of a deep learning model based on a PyTorch framework.
In specific implementation, the invention can adopt an asynchronous random gradient descent method to carry out deep learning training, the learning target is to train an optimal face recognition model suitable for face recognition, each participating mechanism can negotiate whether to use a pre-training model, and if so, what pre-training model, such as FaceNet, is required to be used.
The implementation subject of the blockchain-based face recognition method shown in fig. 1 may be a blockchain-based face recognition system, and the blockchain-based face recognition system and the corresponding blockchain link points form a participation structure. As can be seen from the flow shown in fig. 1, the face recognition method based on the blockchain in the embodiment of the invention can protect the user privacy data while improving the accuracy of face recognition.
Fig. 3 is a flow chart of creating an optimal face recognition model in an embodiment of the invention. Fig. 4 is a schematic diagram of creating an optimal face recognition model according to an embodiment of the present invention. 3-4, uploading face recognition model parameters to a blockchain network to enable the blockchain network to generate global model parameters according to the face recognition model parameters; the creating the optimal face recognition model according to the global model parameters specifically comprises the following steps:
the following iterative processing is performed:
s201: and uploading the face recognition model parameters to corresponding block chain link points in the block chain network for broadcasting, so that one block chain node of the block chain network generates global model parameters according to the face recognition model parameters from each block chain node and broadcasts.
Each blockchain node in the blockchain network is located in one participating mechanism and corresponds to one blockchain-based face recognition system. The N participating institutions upload local face recognition model parameters via internal blockchain nodes. As shown in fig. 4, the face recognition model parameters are uploaded, i.e., the local updates are uploaded.
All participating institutions start blockchain federation networks locally and deploy intelligent contracts for parameter aggregation in the networks, respectively agg And intelligent Contract for model evaluation eval . The blockchain coalition network allows only authenticated participating institutions to access the network, such as building a blockchain coalition chain based on ethernet, as shown in fig. 2. Intelligent contracts for parameter aggregation and model evaluation may be written using the solubility language and deployed in a federation chain.
In one embodiment, generating global model parameters by one of the blockchain nodes of the blockchain network based on face recognition model parameters from each of the blockchain nodes includes:
one of the blockchain nodes of the blockchain network is contacted by the intelligent Contract agg And finishing parameter aggregation, screening face recognition model parameters according to a preset similarity threshold and the similarity between the face recognition model parameters, generating global model parameters according to the screened face recognition model parameters, and broadcasting.
In specific implementation, the intelligent Contract agg And calculating the similarity between the face recognition model parameters uploaded by any two participating institutions according to the cosine similarity or the Jacquard similarity algorithm. And if the similarity is greater than the similarity threshold, deleting one face recognition model parameter corresponding to the similarity. And generating global model parameters according to the deleted face recognition model parameters and broadcasting. The global model parameter may be an average value of a sum of the deleted face recognition model parameters.
S202: and acquiring global model parameters, and creating a face recognition model according to the global model parameters.
As shown in fig. 4, global model parameters are acquired, i.e., global updates are downloaded. The downloading of the global update can be realized by initiating a transaction of any type to the contract address, and calling the API query block data provided by the Ethernet to acquire the global update.
S203: and determining the accuracy of the face recognition model according to the local face test data.
After executing S203, further comprising: uploading the accuracy of the face recognition model to corresponding block link points in the block chain network for broadcasting so that one block chain node of the block chain network passes through an intelligent Contract eval And determining the packed rewards and the blockchain packing nodes according to the accuracy of the face recognition model from each blockchain node, and sending the packed rewards to the blockchain packing nodes, wherein the blockchain packing nodes obtain the packed rewards.
In specific implementation, the intelligent Contract eval Generating an evaluation value of the face recognition model according to the accuracy of the face recognition model, and determining a packaging rewarding report of the jth iteration according to the evaluation value j And taking the blockchain node corresponding to the maximum value of the evaluation value as a blockchain packaging node, wherein the face recognition model parameter uploaded by the blockchain packaging node is the optimal face recognition model parameter.
S204: judging whether the accuracy is larger than a preset accuracy.
S205: and when the accuracy is larger than the preset accuracy, determining the face recognition model as the optimal face recognition model.
S206: and when the accuracy is smaller than or equal to the preset accuracy, updating the global model parameters according to the local face training data, determining the updated global model parameters as face recognition model parameters, and continuing to execute iterative processing.
Conventional deep learning requires that all training data be collected to the same group of servers or data centers for storage, and thus privacy data leakage may occur. As shown in FIG. 2, the invention does not need to upload local training data, but only needs to participate in the mechanism to keep the local face training data on the local equipment.
The local face training data is historical face image data of a user, which is acquired by each participating mechanism by using image acquisition equipment such as a mobile phone and the like, and comprises the historical face image data and a historical face actual result.
Updating global model parameters according to the local face training data comprises: inputting the historical face image data into a face recognition model to obtain a historical face recognition result; determining precision loss according to the historical face actual result and the historical face recognition result; and updating the global model parameters according to the precision loss and the preset learning rate.
The global model parameters and the face recognition model parameters comprise parameter weights and parameter biases, and the global model parameters can be updated through a random gradient descent algorithm. When the random gradient descent algorithm is used for training local data, in order to reduce the volume of model parameters and improve communication efficiency, the invention uses an FP16 floating point semi-precision tensor when the gradient propagates forward and backward, and uses an FP32 floating point single-precision tensor in a model updating stage. Taking the updating parameter weight as an example, the global model parameter weight is updated by the following formula:
wherein w is j+1 For the updated global model parameters, i.e. the face recognition model parameter weights uploaded to the blockchain network in the j+1th iteration, w j And the global model parameter in the jth iteration is a learning rate, and alpha is a learning rate. Global model parameters and face recognition model parameters use FP16 floating point semi-precision tensor
The specific flow of the embodiment of the invention is as follows:
1. the face recognition system uploads the face recognition model parameters to corresponding block link points in the block chain network for broadcasting. The face recognition system and the corresponding block link points form a participation mechanism.
2. One of the blockchain nodes of the blockchain network is contacted by the intelligent Contract agg Completing parameter aggregation, intelligent Contract agg And calculating the similarity between the face recognition model parameters uploaded by any two participating institutions according to the cosine similarity or the Jacquard similarity algorithm. If the similarity is greater than the similarity threshold, the pair of similarities is deletedAnd generating global model parameters according to one face recognition model parameter after deletion and broadcasting the global model parameters.
3. The face recognition system acquires the global model parameters, and creates a face recognition model according to the global model parameters.
4. The face recognition system determines the accuracy of the face recognition model according to the local face test data.
5. The face recognition system uploads the accuracy of the face recognition model to the corresponding block link points in the block chain network for broadcasting.
6. One of the blockchain nodes of the blockchain network is contacted by the intelligent Contract eval And determining the packed rewards and the blockchain packing nodes according to the accuracy of the face recognition model from each blockchain node, and sending the packed rewards to the blockchain packing nodes, wherein the blockchain packing nodes obtain the packed rewards.
7. The face recognition system judges whether the accuracy is larger than a preset accuracy. When the accuracy is larger than the preset accuracy, determining the face recognition model as an optimal face recognition model, otherwise, inputting the historical face image data into the face recognition model to obtain a historical face recognition result; determining precision loss according to the historical face actual result and the historical face recognition result; and (3) updating global model parameters according to the precision loss and the preset learning rate, determining the updated global model parameters as face recognition model parameters, and returning to the step (1).
8. Current facial image data of the user is acquired.
9. And inputting the current facial image data into an optimal face recognition model created based on the local face training data and face recognition model parameters of block chain nodes in the block chain network, and obtaining a current face recognition result.
In summary, the invention has the following beneficial effects:
1. the face recognition method based on the blockchain technology trains the face image data to construct face recognition, removes a centralized parameter server in collaborative deep learning, avoids the disfigurement of the centralized parameter server, prevents the centralized parameter server from stealing the face image data of the user of the collaborative deep learning participation mechanism, and protects the privacy of the face image data and the privacy information of the user.
2. According to the invention, global model parameters are generated through intelligent contracts in the blockchain, the face recognition model parameters uploaded by the participating mechanisms are evaluated, the participating mechanisms uploading the optimal face recognition model parameters are used as nodes of the next round of packing blocks, the participating mechanisms are stimulated to share the optimal face recognition model parameters, the accuracy and generalization of the finally constructed face recognition model are indirectly improved, and the accuracy of face recognition by using the model is improved.
3. According to the invention, the similarity algorithm is adopted to detect the participation mechanism uploading similar face recognition model parameters, so that malicious parameters uploaded by the participation mechanism are stopped to a certain extent, and the accuracy and generalization of the face recognition model are further improved.
4. According to the deep learning architecture constructed based on the block chain, the participation mechanism only needs to send and receive global model parameters to the local, so that the communication efficiency is improved. The participation mechanism performs local training, uses the FP16 floating point semi-precision tensor when the gradient is propagated forward and backward, uses the FP32 floating point single-precision tensor in the model updating stage, reduces the volume of model parameters, improves the propagation speed of a blockchain network, further accelerates the training speed of facial image data, and accelerates the construction speed of a face recognition model.
5. The invention indirectly realizes the 'sharing' of the facial image data of the user between the participating institutions based on the distributed blockchain technology, so that the user does not need to collect the facial data for multiple times when using the face recognition system, the convenience of use is improved, the application of the face recognition system in daily life is further promoted, and the privacy of the facial image data of the local user of the participating institutions is also protected.
Based on the same inventive concept, the embodiment of the invention also provides a face recognition system based on the block chain, and because the principle of solving the problem of the system is similar to that of the face recognition method based on the block chain, the implementation of the system can refer to the implementation of the method, and the repetition is omitted.
Fig. 5 is a block diagram of a block chain based face recognition system in an embodiment of the present invention. As shown in fig. 5, the blockchain-based face recognition system includes:
an acquisition unit configured to acquire current face image data of a user;
the face recognition result unit is used for inputting the current face image data into the optimal face recognition model to obtain a current face recognition result; the step of creating the optimal face recognition model comprises the following steps: uploading the face recognition model parameters to a blockchain network so that the blockchain network generates global model parameters according to the face recognition model parameters; and creating an optimal face recognition model according to the global model parameters.
In one embodiment, the method further comprises:
the optimal face recognition model creation unit is used for executing the following iterative processing:
uploading face recognition model parameters to corresponding block chain link points in a block chain network for broadcasting, so that one block chain node of the block chain network generates global model parameters according to the face recognition model parameters from each block chain node and broadcasts the global model parameters;
acquiring global model parameters, and creating a face recognition model according to the global model parameters;
determining the accuracy of the face recognition model according to the local face test data;
and when the accuracy is larger than the preset accuracy, determining the face recognition model as an optimal face recognition model, otherwise, updating global model parameters according to the local face training data, determining the updated global model parameters as the face recognition model parameters, and continuing to execute iterative processing.
In one embodiment, the optimal face recognition model creation unit is specifically configured to:
and screening face recognition model parameters according to the similarity between a preset similarity threshold and each face recognition model parameter by one block chain link point of the block chain network, generating global model parameters according to the screened face recognition model parameters, and broadcasting.
In one embodiment, the local face training data includes historical face image data and historical face actual results;
the optimal face recognition model creation unit is specifically configured to:
inputting the historical face image data into a face recognition model to obtain a historical face recognition result;
determining precision loss according to the historical face actual result and the historical face recognition result;
and updating the global model parameters according to the precision loss and the preset learning rate.
In one embodiment, the method further comprises:
and the rewarding unit is used for uploading the accuracy of the face recognition model to the corresponding block link points in the block chain network for broadcasting, so that one block chain node of the block chain network determines the packed rewards and the block chain packing nodes according to the accuracy of the face recognition model from each block chain node and sends the packed rewards to the block chain packing nodes.
In summary, the face recognition system based on the blockchain can protect user privacy data while improving the face recognition accuracy.
The embodiment of the invention also provides a concrete implementation mode of the computer equipment capable of realizing all the steps in the face recognition method based on the block chain. Fig. 6 is a block diagram of a computer device according to an embodiment of the present invention, and referring to fig. 6, the computer device specifically includes:
a processor (processor) 601 and a memory (memory) 602.
The processor 601 is configured to invoke a computer program in the memory 602, where the processor executes the computer program to implement all the steps in the blockchain-based face recognition method in the above embodiment, for example, the processor executes the computer program to implement the following steps:
acquiring current facial image data of a user;
inputting the current facial image data into an optimal face recognition model to obtain a current face recognition result; the step of creating the optimal face recognition model comprises the following steps: uploading the face recognition model parameters to a blockchain network so that the blockchain network generates global model parameters according to the face recognition model parameters; and creating an optimal face recognition model according to the global model parameters.
In summary, the computer device of the embodiment of the invention can protect the privacy data of the user while improving the accuracy of face recognition.
The embodiment of the present invention also provides a computer readable storage medium capable of implementing all the steps in the blockchain-based face recognition method in the above embodiment, where the computer readable storage medium stores a computer program, and when the computer program is executed by a processor, the computer program implements all the steps in the blockchain-based face recognition method in the above embodiment, for example, the processor implements the following steps when executing the computer program:
acquiring current facial image data of a user;
inputting the current facial image data into an optimal face recognition model to obtain a current face recognition result; the step of creating the optimal face recognition model comprises the following steps: uploading the face recognition model parameters to a blockchain network so that the blockchain network generates global model parameters according to the face recognition model parameters; and creating an optimal face recognition model according to the global model parameters.
In summary, the computer readable storage medium of the embodiment of the invention can protect user privacy data while improving the accuracy of face recognition.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
Those of skill in the art will further appreciate that the various illustrative logical blocks (illustrative logical block), units, and steps described in connection with the embodiments of the invention may be implemented by electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components (illustrative components), elements, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design requirements of the overall system. Those skilled in the art may implement the described functionality in varying ways for each particular application, but such implementation is not to be understood as beyond the scope of the embodiments of the present invention.
The various illustrative logical blocks, or units, or devices described in the embodiments of the invention may be implemented or performed with a general purpose processor, a digital signal processor, an Application Specific Integrated Circuit (ASIC), a field programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described. A general purpose processor may be a microprocessor, but in the alternative, the general purpose processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, e.g., a digital signal processor and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a digital signal processor core, or any other similar configuration.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may be stored in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. In an example, a storage medium may be coupled to the processor such that the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integral to the processor. The processor and the storage medium may reside in an ASIC, which may reside in a user terminal. In the alternative, the processor and the storage medium may reside as distinct components in a user terminal.
In one or more exemplary designs, the above-described functions of embodiments of the present invention may be implemented in hardware, software, firmware, or any combination of the three. If implemented in software, the functions may be stored on a computer-readable medium or transmitted as one or more instructions or code on the computer-readable medium. Computer readable media includes both computer storage media and communication media that facilitate transfer of computer programs from one place to another. A storage media may be any available media that can be accessed by a general purpose or special purpose computer. For example, such computer-readable media may include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that may be used to carry or store program code in the form of instructions or data structures and other data structures that may be read by a general or special purpose computer, or a general or special purpose processor. Further, any connection is properly termed a computer-readable medium, e.g., if the software is transmitted from a website, server, or other remote source via a coaxial cable, fiber optic cable, twisted pair, digital Subscriber Line (DSL), or wireless such as infrared, radio, and microwave, and is also included in the definition of computer-readable medium. The disks (disks) and disks (disks) include compact disks, laser disks, optical disks, DVDs, floppy disks, and blu-ray discs where disks usually reproduce data magnetically, while disks usually reproduce data optically with lasers. Combinations of the above may also be included within the computer-readable media.

Claims (10)

1. A blockchain-based face recognition method, comprising:
acquiring current facial image data of a user;
inputting the current facial image data into an optimal face recognition model to obtain a current face recognition result; the creating step of the optimal face recognition model comprises the following steps: uploading face recognition model parameters to a blockchain network so that the blockchain network generates global model parameters according to the face recognition model parameters; creating an optimal face recognition model according to the global model parameters;
uploading face recognition model parameters to a blockchain network so that the blockchain network generates global model parameters according to the face recognition model parameters; the creating an optimal face recognition model according to the global model parameters specifically comprises:
the following iterative processing is performed:
uploading face recognition model parameters to corresponding block chain link points in a block chain network for broadcasting, so that one block chain node of the block chain network generates global model parameters according to the face recognition model parameters from each block chain node and broadcasts the global model parameters;
acquiring the global model parameters, and creating a face recognition model according to the global model parameters;
determining the accuracy of the face recognition model according to the local face test data;
and when the accuracy is larger than the preset accuracy, determining that the face recognition model is an optimal face recognition model, otherwise, updating the global model parameters according to the local face training data, determining that the updated global model parameters are face recognition model parameters, and continuing to execute the iterative processing.
2. The blockchain-based face recognition method of claim 1, wherein one of the blockchain nodes of the blockchain network generates global model parameters from face recognition model parameters from each blockchain node includes:
and screening the face recognition model parameters according to a preset similarity threshold and the similarity between each face recognition model parameter by one block chain link point of the block chain network, generating global model parameters according to the screened face recognition model parameters, and broadcasting.
3. The blockchain-based face recognition method of claim 2, wherein the local face training data includes historical face image data and historical face actual results;
updating the global model parameters according to the local face training data comprises:
inputting the historical facial image data into the face recognition model to obtain a historical face recognition result;
determining precision loss according to the historical face actual result and the historical face recognition result;
and updating the global model parameters according to the precision loss and a preset learning rate.
4. The blockchain-based face recognition method of claim 2, further comprising:
uploading the accuracy of the face recognition model to corresponding blockchain link points in a blockchain network for broadcasting, so that one blockchain node of the blockchain network determines packing rewards and blockchain packing nodes according to the accuracy of the face recognition model from each blockchain node, and sends the packing rewards to the blockchain packing nodes.
5. A blockchain-based face recognition system, comprising:
an acquisition unit configured to acquire current face image data of a user;
the face recognition result unit is used for inputting the current face image data into an optimal face recognition model to obtain a current face recognition result; the creating step of the optimal face recognition model comprises the following steps: uploading face recognition model parameters to a blockchain network so that the blockchain network generates global model parameters according to the face recognition model parameters; creating an optimal face recognition model according to the global model parameters;
the face recognition system further includes:
the optimal face recognition model creation unit is used for executing the following iterative processing:
uploading face recognition model parameters to corresponding block chain link points in a block chain network for broadcasting, so that one block chain node of the block chain network generates global model parameters according to the face recognition model parameters from each block chain node and broadcasts the global model parameters;
acquiring the global model parameters, and creating a face recognition model according to the global model parameters;
determining the accuracy of the face recognition model according to the local face test data;
and when the accuracy is larger than the preset accuracy, determining that the face recognition model is an optimal face recognition model, otherwise, updating the global model parameters according to the local face training data, determining that the updated global model parameters are face recognition model parameters, and continuing to execute the iterative processing.
6. The blockchain-based face recognition system of claim 5, wherein the optimal face recognition model creation unit is specifically configured to:
and screening the face recognition model parameters according to a preset similarity threshold and the similarity between each face recognition model parameter by one block chain link point of the block chain network, generating global model parameters according to the screened face recognition model parameters, and broadcasting.
7. The blockchain-based face recognition system of claim 6, wherein the local face training data includes historical face image data and historical face actual results;
the optimal face recognition model creation unit is specifically configured to:
inputting the historical facial image data into the face recognition model to obtain a historical face recognition result;
determining precision loss according to the historical face actual result and the historical face recognition result;
and updating the global model parameters according to the precision loss and a preset learning rate.
8. The blockchain-based face recognition system of claim 6, further comprising:
and the rewarding unit is used for uploading the accuracy of the face recognition model to corresponding block chain link points in the block chain network for broadcasting, so that one block chain node of the block chain network determines packing rewards and block chain packing nodes according to the accuracy of the face recognition model from each block chain node and sends the packing rewards to the block chain packing nodes.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and running on the processor, characterized in that the processor implements the steps of the blockchain-based face recognition method of any of claims 1 to 4 when the computer program is executed.
10. A computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements the steps of the blockchain-based face recognition method of any of claims 1 to 4.
CN202010762414.3A 2020-07-31 2020-07-31 Face recognition method and system based on block chain Active CN111950416B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010762414.3A CN111950416B (en) 2020-07-31 2020-07-31 Face recognition method and system based on block chain

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010762414.3A CN111950416B (en) 2020-07-31 2020-07-31 Face recognition method and system based on block chain

Publications (2)

Publication Number Publication Date
CN111950416A CN111950416A (en) 2020-11-17
CN111950416B true CN111950416B (en) 2023-08-29

Family

ID=73339134

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010762414.3A Active CN111950416B (en) 2020-07-31 2020-07-31 Face recognition method and system based on block chain

Country Status (1)

Country Link
CN (1) CN111950416B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR102520513B1 (en) * 2021-11-16 2023-04-11 주식회사 딥이티 Apparatus and method for face recognition using user terminal

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110197128A (en) * 2019-05-08 2019-09-03 华南理工大学 The recognition of face architecture design method planned as a whole based on edge calculations and cloud
CN111062339A (en) * 2019-12-19 2020-04-24 广州广大通电子科技有限公司 Face recognition method, device, equipment and storage medium based on block chain

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110197128A (en) * 2019-05-08 2019-09-03 华南理工大学 The recognition of face architecture design method planned as a whole based on edge calculations and cloud
CN111062339A (en) * 2019-12-19 2020-04-24 广州广大通电子科技有限公司 Face recognition method, device, equipment and storage medium based on block chain

Also Published As

Publication number Publication date
CN111950416A (en) 2020-11-17

Similar Documents

Publication Publication Date Title
US20170034217A1 (en) Authorization of access to a data resource in addition to specific actions to be performed on the data resource based on an authorized context enforced by a use policy
CN110291514A (en) Intelligent contract data library is traversed by logical mappings
CN113792871A (en) Neural network training method, target identification method, device and electronic equipment
CN106600243A (en) Mobile payment method and system based on mixed mode
CN104054078A (en) Method for providing a set of services of a first subset of a social network to a user of a second subset of said social network
CN109729044A (en) A kind of general internet data acquisition is counter to climb system and method
WO2022199480A1 (en) Multi-party collaborative model updating method, device, and system for realizing privacy protection
Gupta et al. Learner’s dilemma: IoT devices training strategies in collaborative deep learning
CN111368984B (en) Method and device for league learning and league learning system
CN111950416B (en) Face recognition method and system based on block chain
CN112488163A (en) Abnormal account identification method and device, computer equipment and storage medium
US20180337989A1 (en) System and method for synchronizing data between computer devices
CN112101577B (en) XGboost-based cross-sample federal learning and testing method, system, device and medium
CN111340574B (en) Risk user identification method and device and electronic equipment
JP2023535851A (en) METHOD, DEVICE, TERMINAL DEVICE, AND STORAGE MEDIUM FOR DATA PROCESSING MODEL BY PRIVACY PROTECTION
CN117151859A (en) Model training and evaluating method, related device, equipment, medium and product
Upadhyay et al. Feature extraction approach for speaker verification to support healthcare system using blockchain security for data privacy
CN109388747A (en) The method and apparatus of the confidence level of user in a kind of acquisition network
CN107346517A (en) User-interaction parameter acquisition methods and acquisition device in customer relationship network
CN113033209B (en) Text relation extraction method and device, storage medium and computer equipment
CN115687526A (en) Seismic data model sharing method based on block chain and federal learning
CN114970670A (en) Model fairness assessment method and device
CN112819177B (en) Personalized privacy protection learning method, device and equipment
CN111914065B (en) Short message content verification method, device, computer system and computer readable medium
CN117272370B (en) Method, system, electronic equipment and medium for recommending privacy protection of next interest point

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant