43results about How to "Guarantee privacy and security" patented technology

The application discloses a screen projection method, device and system, electronic equipment and a storage medium. The screen projection method comprises the following steps: receiving a broadcast signal; analyzing the broadcast signal to obtain screen projection information; judging whether to display a screen projection request prompt based on the screen projection information; if it is judged to display the screen projection request prompt, the screen projection request prompt is displayed; and in response to a confirmation operation of the user agreeing to the screen projection request prompt, a screen projection operation is performed. The application triggers the screen projection operation by the screen projection end, specifically, the broadcast signal sent by the screen projection end is received, and the screen projection operation is performed according to the screen projection information carried in the broadcast signal, so that the user does not need to search for the screen projection end on the screen projection end, the operation is convenient, the error rate of the screen projection connection is reduced, and the user experience is improved.

The application discloses a kind of emotion evaluation and intervention method and system based on multi-modal data.Through terminal synchronous acquisition voice, face, screen tremor and so on multi-modal data;Reaction mode vector is generated based on context information and data verification fusion;The composite emotional state of user is determined using emotional emergence model.At the same time, personalized emotional causal diagram is constructed and updated, and the enhanced state vector is generated by extracting key causal path features.The vector is input into the intervention agent to output intervention actions;After execution, the emotional state is evaluated again and the reward is calculated, and the system model is updated online.The application realizes accurate emotion evaluation and closed-loop adaptive intervention that strikes the root.

The present application relates to the technical field of vehicle pseudonym replacement, and particularly relates to a vehicle pseudonym replacement method based on artificial intelligence, which comprises the following steps: determining whether pseudonym replacement is needed according to a replacement time interval value or performing secondary determination based on a quantity ratio reference value and a communication characteristic value; determining a target area type according to a to-be-replaced vehicle distribution reference value and a replacement allowance, and determining whether to directly perform pseudonym replacement or perform combined analysis according to the target area type; in the combined analysis, determining a related vehicle combination based on a position synchronization degree or an interaction reference value according to a vehicle analysis coefficient, determining whether to perform sequential setting of a virtual vehicle based on a quantity difference value or setting a virtual vehicle based on a historical attack frequency of a sub-area, and performing pseudonym replacement on to-be-replaced vehicles and virtual vehicles in the related vehicle combination. The present application can reduce the risk of vehicle location privacy leakage.

The application discloses a kind of stealthy photographing behavior detection method, device, electronic equipment and storage medium, the present application first collects the pedestrian image in target area, then, through machine recognition technology, to identify the arm posture of each pedestrian in pedestrian image, and determine the stealthy photographing suspect object according to arm posture;Then, again through the hand-held object type of identifying the stealthy photographing suspect object, and the dress type of pedestrian in its corresponding stealthy photographing image area, to finally determine whether the stealthy photographing suspect object exists stealthy photographing behavior;Finally, after determining that there is stealthy photographing behavior, then the interception of stealthy photographing image can be carried out, and sent to monitoring terminal, to realize the retention of evidence and secondary detection of stealthy photographing behavior;Therefore, the present application uses machine recognition technology to realize stealthy photographing behavior detection, compared with artificial patrol, efficiency is higher, detection accuracy is higher, effectiveness is stronger, can produce greater deterrent to lawbreaker, can fully guarantee the privacy security of passenger.

This invention belongs to the field of identity authentication in the Internet of Vehicles (IoV), specifically relating to an anonymous batch authentication method for IoV commuting environments. The invention includes: constructing an IoV system comprising a Trusted Authority (TA), roadside units (Roadside Units), and vehicles; publishing and initializing system parameters; vehicles and Roadside Units pre-registering and verifying their identities with the TA; the TA verifies the identity, calculates and returns key parameters and a pseudonym; vehicles calculate their complete private and public keys based on the returned parameters and send a verification message to the TA; the TA verifies the identity and creates a vehicle authentication schedule; commuting vehicles initiate authentication to Roadside Units (Roadside Units) on the commuting route; the Roadside Units receive the message broadcast by the vehicles, verify the legitimacy of the vehicle requests using the vehicle authentication schedule, and complete the authentication. This invention ensures anonymous and secure communication among entities in the IoV system and reduces computational and communication overhead, thereby improving authentication efficiency.

The application relates to a user data storage management method and an embedded device. The method is applied to the embedded device, user conversation data is acquired, local end information extraction is performed on the user conversation data based on preset extraction conditions, current user key data is obtained, original user key data stored in a first storage unit is acquired according to the current user key data, hierarchical storage judgment is performed on the current user key data and the original user key data, first storage data meeting first preset storage conditions is stored in the first storage unit, second storage data meeting second preset storage conditions is stored in a second storage unit, a user search instruction is acquired, the first storage unit and the second storage unit are searched according to the user search instruction and a preset search algorithm, target data is obtained, user privacy data leakage is avoided, user privacy safety is guaranteed from an architecture level, and the reliability of storage management in a resource-limited platform is improved.

The application relates to the technical field of artificial intelligence, and discloses a large model expert library construction based on a knowledge graph, which realizes real-time acquisition of multi-source data, marks an update point through a hash fingerprint algorithm and semantic similarity calculation, and stores the update point in a message queue in combination with metadata determination; a low-dimensional semantic vector is generated by extracting features and constructing a fusion network, and cross-modal mapping is established; an attribute graph is built with the semantic vector as a node, and the attribute graph contains basic facts, reasoning rules and time sequence edges; a user request is parsed as a graph traversal task, and a result is returned through counterfactual reasoning; sensitive data and relationships are encrypted and indexed; source field graph features are migrated to optimize target field graph, and the construction of the expert library is accelerated. The application realizes efficient data processing, semantic correlation analysis, safe query and rapid construction of the expert library, and improves data utilization and knowledge reasoning efficiency.

The invention provides a multi-layer illusion special effect resource generation method and device, equipment and a storage medium, and relates to the technical field of image processing. The method comprises the following steps: displaying a phantom multi-layer configuration interface; receiving a resource packet corresponding to each layer; carrying out blank area identification on the sequence frame image and the default static image of each image layer, calculating a uniform cuttable area, and carrying out cutting processing to obtain a target sequence frame image and a target static image of each image layer; based on a dragging operation of a user or an offset obtained by form input, adjusting the offset of the offset layer relative to the fixed layer, and outputting a target offset; and converting the target sequence frame images of each image layer into target animations in batches, exporting the target animations, generating a JSON configuration file, compressing and packaging the target animations, the target static images and the JSON configuration file, and generating a illusion special effect resource packet. And independent management of multi-layer resources, automatic identification and cutting of blank areas and accurate offset control of the layers are realized.

The invention provides an AI video monitoring platform based on an RISC-V framework. The AI video monitoring platform comprises eight levels of frameworks including an equipment perception layer, an edge computing layer, a data access layer, a data processing layer, a data storage layer, an algorithm engine layer, a platform application layer and an industry application layer. The invention belongs to the technical field of monitoring platforms, particularly relates to an AI video monitoring platform of an RISC-V architecture, and aims to at least solve the problems that traditional security depends on manual screen staring and inspection, abnormal event recognition lags behind and the missing report rate is high, and the existing system is high in authorization cost, insufficient in edge scene energy efficiency ratio and poor in adaptability. The risks that the data transmission delay is high, real-time early warning cannot be supported and privacy leakage exists in original video cloud storage due to a front end-cloud framework are avoided; the method avoids the phenomena that different brand devices and platform protocols are not uniform, the compatibility is poor, the deployment and maintenance cost is high, and the expansion capability is weak, and overcomes the defects that a general algorithm is insufficient in precision, lacks an industry exclusive algorithm, and cannot cover the multi-scene safety protection requirement.

The invention relates to a smart home and smart property agent cooperation system and method, the system comprises a home agent, a property agent and a community agent, and the home agent is used for processing multi-modal input of a user according to a personalized decision model and generating event information according to a privacy strategy. And the property agent is used for performing task reasoning according to the event information and a preset knowledge base and generating a corresponding task instruction. And the community agent is used for executing the task instruction and feeding back an execution result to the property agent. Wherein the home agent, the property agent and the community agent perform data interaction through a preset communication architecture. The problem of how to improve the business processing efficiency while the intelligent agents of the intelligent home and the intelligent property management system cooperate in the related technology is solved.

The invention discloses a multi-person shared health detection device capable of accurately recognizing user identity based on sound. The device comprises a microphone array, a physiological parameter sensor, a processor and a memory. The method comprises the following steps of: performing front-end processing on a collected voice signal to obtain an effective voice segment; extracting a voiceprint embedding vector by using a deep neural network; comparing the vector with a local voiceprint template library to obtain a highest score and candidate identities; carrying out living body and playback attack authenticity discrimination by analyzing multi-dimensional acoustic features and utilizing a classification model; and binding and associating the identity label with the health data collected by the sensor when the identity label is judged to be a real living body and the similarity score meets a judgment condition. The method has the advantages of high noise resistance, attack resistance and end-side low delay, supports offline work and model increment self-adaption, is suitable for families, communities, medical institutions and other scenes, and effectively improves the use convenience and data security of shared health equipment.

The application relates to the field of video applications, and provides a video desensitization method and device, electronic equipment and a computer program product. The method comprises the following steps: acquiring video data, identifying a human hand region based on the video data, obtaining N gesture recognition results and M hand contact article recognition results, determining a sensitive region and a target sensitive level according to the N gesture recognition results and the M hand contact article recognition results, and performing desensitization processing on the video data according to the sensitive region and the target sensitive level. The video desensitization method provided in the embodiment of the application can solve the technical problem that a hand contact article or hand operation of a photographed person cannot be desensitized, personal information of the photographed person is easily leaked, and the personal information safety of the photographed person is improved.

The utility model belongs to the technical field of charging treasure, concretely relates to magnetic formula hidden detector charging treasure, including upper casing and lower casing, the upper casing is connected with lower casing, the upper casing top assembly has the mounting panel, the mounting panel top surface fixedly connected with the top cover, the mounting panel top surface assembly has the circuit board, the upper casing inner chamber assembly has the battery group, the battery group is connected with circuit board electricity, the circuit board electricity is connected with the peep detector on, the top cover is equipped with the through -hole with the peep detector corresponding, the upper casing side wall assembly has the magnetic sheet. The upper casing inner wall fixedly connected with the bearing plate, power button and display screen are installed to the bearing plate side wall. The utility model can integrate the peep detector with the charging treasure together, is convenient for the detection to the hidden camera when moving into the hotel.

This application relates to a system and method for reducing the number of parameters transmitted in federated learning across food sectors, belonging to the field of artificial intelligence technology. The system includes: a non-holonomic neural network construction module for constructing a non-holonomic neural network model with fewer parameters than a full neural network; a convolutional kernel generation and training module for generating an updated non-holonomic neural network model using a convolutional kernel generation function and performing federated learning training on the non-holonomic neural network model; a module upload and combine modules, which randomly assign numbers to each client via a server and instruct them to upload one module from the updated non-holonomic local model, with the server then reassembling a new global model; and a model evaluation module for testing the global model using a test dataset after federated learning training and evaluating the model performance based on the test results. This system addresses the problem of "food data silos" while ensuring efficient communication in federated learning.

The invention discloses a data desensitization processing method based on AI. The data desensitization processing method comprises the following steps that 1, multi-modal data preprocessing and sensitive field intelligent identification are carried out; step 2, intelligent matching of a desensitization strategy; 3, performing multi-dimensional verification and self-adaptive optimization on the desensitization effect; according to the method, the sensitive field recognition precision is high, the efficiency is excellent, a recognition framework with the rule engine and the fine-tuning AI model fused is adopted, the rule engine achieves rapid recognition of the high-confidence sensitive field, and the recognition efficiency is improved; the fine-tuning AI model is trained through an industry exclusive sample set, has good industry adaptability, improves the recognition precision in a complex scene, eliminates recognition conflicts in combination with a D-S evidence theory, and has the overall recognition accuracy not lower than 96%, which is significantly superior to that of an existing single recognition method.

The application discloses a psychological intervention method and system based on dynamic intention routing and thought chain reasoning, and belongs to the technical field of artificial intelligence and mental health. The method comprises the following steps: performing feature scanning on user input through a double-track detection mechanism composed of a semantic similarity model and a sensitive word matching algorithm, and generating an emotional tendency score and a crisis risk weight; comparing the scanning result with a preset threshold, and shunting to a red early warning, yellow intervention or green general path; triggering a three-level early warning action for the red path; starting a thought chain reasoning engine for the yellow path, sequentially calling multiple logic operators, and generating a step-by-step traceable intervention strategy; calling a general dialogue engine for the green path; and dynamically adjusting the guidance intensity and the level of empathy based on user feedback. The application realizes millisecond-level recognition and forced interruption of high-risk intentions, ensures the professionalism and interpretability of the intervention process through structured reasoning, and significantly improves the safety, professionalism and response efficiency of the psychological intervention system.

This invention relates to the field of medical information security technology, and discloses a blockchain-based method and system for auditing and detecting unauthorized access to medical data. The method includes: structurally modeling medical data access behavior to construct access behavior feature vectors; performing multi-source fusion risk assessment based on an access control policy model and a behavior learning model to obtain a comprehensive risk value; executing a graded processing strategy according to the risk level and generating corresponding audit logs; structurally organizing the audit logs and constructing a Merkle tree to generate data fingerprints; writing the data fingerprints and summary information into the main chain, storing the complete audit logs in a side chain or off-chain storage, and achieving consistency verification through Merkle proofs; using smart contracts to verify the audit data and automatically execute access control responses, while simultaneously feeding back the audit results for model updates. This invention enables real-time detection, graded processing, and trusted auditing of medical data access behavior.

The invention relates to the technical field of Internet education data processing and network communication, and discloses a teaching resource sharing decision optimization method based on a hierarchical federated incentive mechanism, and the method specifically comprises the steps: constructing a hierarchical heterogeneous federated topology network based on resource entropy, determining the storage hierarchy and routing strategy of teaching resources through the calculation result of the resource entropy, and carrying out the decision optimization of the teaching resources. Establishing a multi-dimensional reputation weighted right quantification model, and calculating comprehensive reputation weights of participating nodes; and executing a resource scheduling decision based on the dynamic Stackelberg game, solving a Nash equilibrium solution by using a distributed iterative algorithm, determining an optimal price vector and a resource allocation matrix, monitoring a game equilibrium drift rate to implement dynamic topology reconstruction, and triggering topology splitting or fusion operation according to the Nash equilibrium drift rate and a global utility variance. According to the method, privacy and efficiency are balanced through resource entropy hierarchical storage, pricing is optimized by using a reputation weighted game, and self-adaptive stability of a network architecture is guaranteed by means of topology reconstruction based on a drift rate.

The invention relates to the technical field of cloud edge collaboration, in particular to an industrial equipment computing task dynamic allocation method based on cloud edge collaboration, which comprises the following steps of: constructing a computing task global allocation model containing a decentralized privacy protection mechanism, decomposing the computing task global allocation model into a plurality of equipment terminal optimization problems and a cloud global aggregation problem; according to the method, through a mechanism of iteratively and interactively updating global coordination variables and local intermediate variables, whole-network computing power resource collaborative optimization is realized on the premise of not collecting underlying data, a convergence residual error is used as a feedback index to dynamically adjust an allocation strategy, it is ensured that a finally generated task allocation scheme is mathematically converged to a globally optimal solution, and the task allocation efficiency is improved. The network bandwidth congestion pressure caused by real-time uploading of mass equipment state data is reduced, task processing delay caused by overweight calculation load of a single-point data center is avoided, and the response speed of an industrial internet system when the industrial internet system deals with large-scale concurrent requests is increased.

The application is suitable for the technical field of medical big data quality management, and provides a data quality control and automatic repair method and system. The method comprises the following steps: performing federated desensitization slicing on multi-source medical data; performing double time sequence anchoring to obtain a double time sequence offset; constructing a time sequence-entity double-dimensional dynamic constraint graph based on the offset, and calculating an abnormal confidence to detect abnormalities; performing distributed causal consistency verification on abnormal data, and then performing minimum cost automatic repair; writing the repaired data back to the source system, and performing online incremental update on the constraint graph based on the repaired data. Through the closed loop of detection-verification-repair-write-update, the application realizes dynamic self-adaptive optimization of rules, effectively avoids secondary abnormalities introduced by repair, and improves data quality while ensuring data privacy and time sequence accuracy.

The one or more embodiments of the specification provide a label data management method and device based on a multi-tenant environment, which comprises: a platform receiving an information query request sent by a first intelligent agent, analyzing user information of a target user carried therein, and determining an accessible label set thereof; then searching a label database according to the user information, obtaining label information corresponding to the accessible label of the user, the label information being generated based on historical dialogue data of a second intelligent agent and the target user; and finally returning the accessible label and the corresponding label information to the first intelligent agent to support the first intelligent agent to generate personalized dialogue content.

The invention discloses a graph neural network training method, system and device for multi-label graph classification and a medium, and belongs to the technical field of graph neural networks, and the method comprises the steps: constructing a local private graph data set of a client in federated learning; constructing a local model according to the local private graph data set, and performing fusion processing on the local model to obtain local model parameters; aggregating the local model parameters of the client to obtain a global model; and classifying the multi-label graph according to the global model. The problems that in the prior art, feature learning is not comprehensive, the universality of the graph neural network is insufficient, and a label cannot guide training of the graph neural network in a targeted mode are solved.

This application discloses a privacy-preserving feature extraction method and system applicable to data base generation, relating to the field of IoT privacy protection technology. The method includes: operating in a first power consumption mode and with a first computational precision to acquire environmental perception data and obtain a complexity level; if the complexity level exceeds a first preset threshold, waking up a second feature extraction unit to extract first structured feature data and uploading it to a cloud data base; if the confidence score of the first structured feature data is lower than a second preset threshold, extracting local area data and uploading it to the cloud data base; performing refined identification on the local area data to generate second structured feature data and storing it in the cloud data base. This solves the technical problems of insufficient feature extraction accuracy, excessive power consumption, and high bandwidth pressure in existing vehicle-side applications due to edge computing power limitations.

The invention discloses an electric power operator behavior edge identification and AR correction method and system based on a double-flow heterogeneous lightweight network. The method comprises the following steps: acquiring a first visual angle video stream and sensor data of a six-axis inertial measurement unit through intelligent equipment; performing normalization and de-noising processing on the video frame, and segmenting the video frame into short-time video clips according to a time window; running the pre-trained double-flow heterogeneous lightweight network, performing real-time reasoning on the video clip, and outputting a current behavior category label; the behavior confidence degree output by the double-flow heterogeneous lightweight network model is compared with a preset threshold value to judge whether the event is a violation event or not; if the event is identified as a violation event, a deviation correction mechanism is triggered immediately; and encrypting and uploading the violation event log to an enterprise security management platform or other third-party platforms for users to check, count and trace. The method can effectively solve the problems of high behavior identification delay, dependence on cloud, low edge equipment operation efficiency, insufficient multi-behavior identification precision and the like in the prior art.

The invention provides a mobile payment platform user privacy protection method and system, and relates to the technical field of mobile payment, and the method comprises the steps: triggering a data optimization process when the verification of a biological recognition system is successful but the matching score is low, carrying out the calculation in a random access memory of a device, and carrying out the high-intensity encryption of an increment parameter used for optimizing the biological recognition system, thereby achieving the user privacy protection. And the incremental parameters are not stored persistently. And then, instantaneously uploading the encrypted increment parameter to a server in a data stream form through a secure communication channel, and not writing the increment parameter into any local storage medium of the equipment in the whole process. And if the uploading fails, the encrypted increment parameter is destroyed from the random access memory immediately. According to the method, through full life cycle control from data generation, processing, transmission to destruction, non-persistent storage and instantaneous transmission of increment parameters are intensified in an emphasized manner, data are prevented from being traced in a local storage medium, the utilization risk of malicious programs is remarkably reduced, and the privacy security of mobile payment users is effectively guaranteed.

The application relates to the technical field of data security, in particular to a user privacy data security protection method and system for a cloud platform, which solves the technical problem that, in the prior art, a unified privacy protection strategy is adopted for all position information, adaptive differential protection cannot be carried out according to the actual sensitivity of positions, and the technical problems of waste of computing resources and insufficient personalized protection are caused. The method comprises the following steps: acquiring a position time sequence of a target user in multiple time periods; for each time period, performing time sequence feature analysis according to the position time sequence to determine a position influence index of each position; the position influence index is used for representing the influence degree of the position on the moving behavior of the target user; according to the change of the position influence index of each position in the multiple time periods, the data encryption level corresponding to each position is determined, and each position data is stored after being encrypted according to the data encryption level.

The present application relates to the field of self-service settlement machine, especially to a kind of fast identification method and device for palm vein based on multi-level feature index, the method first extracts the palm vein feature of user palm vein image, then in response to the input of user search code, user search code and palm vein feature are input into the database based on multi-level index architecture to carry out hierarchical indexing, obtain index comparison image, then, user palm vein image and index comparison image are compared, and similarity is obtained, finally, whether the similarity is higher than threshold value is judged, if yes, it is determined that user identity authentication is passed.Compared with prior art, the image matching efficiency of the method of the present application is higher, and the indexing of user privacy information such as gender, age, telephone number is not involved throughout, which can better ensure the security of user privacy information.

The application discloses a kind of multi-source data management method and device based on federal mechanism, belong to computer data processing technical field, it includes the local data of each data source node is collected and extracts semantic summary, is formed distributed data source register by consensus verification;Cross-source collaborative analysis request is received, and is parsed as non-cyclic directed task graph;According to task graph, query is executed locally in each node, and collaborative calculation is completed by encrypted shared intermediate result, generates final analysis result and task execution log;Analysis task execution log is generated for each node Dynamic credit score;According to credit score, update node weight to adaptively optimize the node selection of subsequent task.The application adopts the closed-loop management strategy that multi-source heterogeneous data is semantically abstracted, collaborative analysis request is parsed as execution task graph, dynamic credit score is generated in combination with task execution log and is adaptively optimized according to node selection, and cross-domain data collaborative analysis and management of safe, efficient and intelligent can be realized.

The application discloses a cross-platform dark web intelligence dynamic tracking method based on federal learning, relates to the technical field of intelligence tracking, and comprises the following steps: a federal server performs weighted average aggregation on encrypted local model parameter gradients, generates a global model update gradient, distributes the global model update gradient to each federal learning client, each federal learning client updates a local model after decryption, and obtains a new generation of global model; when any federal learning client finds a suspicious target, each federal learning client uses the new generation of global model and a local multi-modal feature vector to perform joint reasoning calculation, generates cross-platform related intelligence, and writes the cross-platform related intelligence into a blockchain as a notarization record. The application improves the accuracy of dark web entity correlation, the model convergence speed and the generalization ability through an integrated intelligence tracking method, and also solves the core pain points in the previous methods, such as multi-source information fragmentation, high privacy leakage risk and difficult intelligence solidification.