243 results about "Crowds" patented technology

The invention relates to the technical field of urban geography, in particular to a crowd movement flow-based multi-scale community structure importance identification method, which comprises the following steps of: constructing a crowd movement space-time network according to a preset geographic space unit and a time interval on the basis of crowd movement starting place and destination data; identifying a dynamic community structure based on a crowd mobile space-time network; performing multi-scale community structure object role and importance identification; and performing pattern classification by using the time sequence importance of the multi-scale community structure object. According to the method, the problems of inaccurate dynamic community structure recognition and lack of multi-scale object contribution decoupling mechanisms are solved.

The invention relates to the field of artificial intelligence, discloses a crowd advertisement accurate putting determination method based on a knowledge graph, and aims to solve the problem of mismatching caused by interest model lag in traditional advertisement putting. The method comprises the following steps: constructing a dynamic evolution type user interest knowledge graph, fusing multi-source behavior flow and static portrait data, and introducing a time decay factor and an event triggering mechanism to realize node weight self-adaptive updating; a short-term interest pulse is accurately captured through a time sequence attention propagation and cross-domain association edge dynamic generation algorithm; and constructing a dual-channel graph neural network inference engine, respectively processing long-term stable and short-term sudden interest paths, and generating a delivery decision through confidence weighted fusion. According to the invention, the timeliness and accuracy of advertisement matching are improved, the mismatching rate is reduced, and the user experience and the self-calibration capability are enhanced.

The invention relates to the technical field of Internet red scenic spot tourist crowd gathering risk monitoring, and discloses an Internet red scenic spot tourist crowd gathering risk intelligent monitoring and early warning system which comprises a central processing unit, a crowd density analysis module and the like. The crowd density analysis module generates basic density distribution based on region division and historical passenger flow data; the environment state acquisition module acquires environment and facility data in real time; the tourist behavior analysis module generates a behavior feature vector through the track and the stay duration; the central processing unit generates a risk early warning instruction through a spatial superposition algorithm; the risk early warning execution module adjusts monitoring parameters according to instructions and feeds back states, and triggers a dynamic adjustment algorithm when abnormity occurs. The system also improves the accuracy and the resource utilization rate through an optimization and adaptation module, achieves the intelligent monitoring and precise early warning of the crowd gathering risk, and guarantees the tourism safety.

The invention provides an elevator multi-mode crowd feature perception and intelligent advertisement putting method and system, and relates to the technical field of intelligent advertisement putting, and the method comprises the steps: collecting multi-mode perception data in an elevator through an edge computing terminal; performing feature decoupling on the data, performing cross-modal semantic alignment, establishing a directed association relationship between modals, and constructing a scene feature map; calculating the topology importance degree of map nodes, screening feature nodes, and extracting context information for semantic coding; mapping the scene semantic code and the advertisement audience semantic code to a two-dimensional coordinate system to construct a semantic matching graph, and extracting an optimal matching path to form a candidate set; predicting a scene evolution trend based on the scene characteristic spectrum evolution trajectory, and calculating an advertisement adaptive score to generate a playing sequence; putting and collecting user interaction data feedback according to the sequence to update the graph structure. According to the invention, accurate crowd feature recognition and advertisement dynamic matching are realized, and the advertisement putting efficiency and the user experience are improved.

The invention relates to the technical field of artificial intelligence, in particular to a video GIS intelligent analysis method based on deep learning, and the method comprises the steps: detecting a target in a video in real time through a predefined target detection model, and generating a space-time mark which comprises a bounding box and category information; thirdly, associating target tracks of different cameras by using a dynamic graph model and a dynamic graph neural network to form a space-time ID and a track chain; then, combining a visual inertial odometer and a geographic information system to calibrate a homography matrix, and mapping the trajectory chain to a geographic coordinate system to obtain space-time trajectory data; then, constructing a neural network model and a trajectory generation model, and respectively predicting crowd density distribution and pedestrian motion trajectories; and finally, carrying out real-time alarm according to the track and the multi-level geo-fencing rule. According to the invention, intelligent analysis of video data is realized, target tracking and alarm capabilities are improved, and the method is widely applicable to the fields of crowd management and safety monitoring.

The invention provides a DMP-based crowd selection method and apparatus, and an electronic device, a data management platform pre-processes multi-source user data, generates user features corresponding to the pre-processed data based on a preset rule, converts the obtained user features into user tags based on a pre-constructed tag system, and sends the user tags to the user management platform; the method comprises the following steps: acquiring a user tag, storing the acquired user tag in a distributed storage mode, displaying the stored user tag through an operation interface, converting a circling operation for a target user tag displayed on the operation interface into a rule tree, and circling a target user corresponding to the target user tag based on the rule tree. And associating the circling result with the identification information of the target user corresponding to the circling result to generate a crowd packet, and expanding the obtained crowd packet by adopting a similar crowd expansion algorithm. According to the invention, the problems of complex rule configuration, insufficient real-time performance, limited expansibility and the like of a DMP-based crowd selection system in the prior art can be solved.

The invention provides a multi-dimensional risk situation awareness method and system based on dynamic people flow modeling, and relates to the technical field of crowd safety management, and the method comprises the steps: obtaining crowd flow data, and generating a distribution situation and a flow situation; constructing a grid flow incidence matrix; convergence and diffusion paths are identified, and risk weights are determined; generating risk probability distribution and grade distribution data; dividing an evacuation area, and planning an evacuation channel; and calculating a safe passing threshold value and a buffer area capacity, and realizing batch evacuation. The risk distribution can be accurately predicted, the evacuation efficiency is improved, and the congestion risk is reduced.

The invention discloses an intelligent question and answer implementation method and system based on a large model and a semantic atlas, and relates to the technical field of supply chain intellectualization, and the method comprises the steps: obtaining a structured distribution data table and an unstructured text data stream, carrying out the data cleaning and normalization processing, and obtaining a structured distribution data table and an unstructured text data stream; constructing a unified knowledge base to perform semantic understanding and knowledge extraction on the unified knowledge base, extracting key entities and semantic relationships through named entity recognition and relationship extraction tasks, and fusing structured distribution data features to construct a preliminary semantic map; and integrating the preliminary semantic map with the spatial dimension features acquired by the data interface and the real-time task execution log of the shop patrol assistant, endowing nodes with space-time attributes through graph neural network training, and outputting a dynamic semantic map integrated with space-time dynamic mode features. According to the method, geofence coordinates and crowd density characteristics are injected by utilizing an Atde API, a store patrol assistant real-time replenishment state log is superposed, and the causal association strength modeling precision is improved through ST-GNN training.

The invention discloses an urban comprehensive park space-time network construction method and system based on crowd trajectory data. The method comprises the following steps: firstly, collecting, cleaning and dividing crowd trajectory data in a comprehensive park range; secondly, scenic spots in the comprehensive park are clustered, the number of optimal clusters is determined through clustering analysis, and all the clusters serve as all functional activity districts; then building a trajectory model through space projection and a shortest path algorithm, recovering a user moving trajectory path according to the crowd trajectory data, generating path node sequence data according to the timestamp of each trajectory point, and finally, taking the centroid of each functional activity area as each node of the space-time network, and combining each time period to obtain the user moving trajectory path. And constructing an undirected space-time network and a directed transfer network. The method has the advantages of efficient data processing, high modeling precision and perfect privacy protection, dynamic monitoring and space optimization analysis of the park people flow can be realized, and reliable data support and decision basis are provided for urban comprehensive park planning management.

The invention discloses an intelligent dynamic guiding method based on a multi-agent system, and relates to the technical field of multi-agent systems, and the method comprises the following steps: S1, constructing an environment macroscopic state abstract tensor, and inputting the environment macroscopic state abstract tensor into an improved QMIX model; s2, extracting an optimal evacuation decision intention tensor; s3, constructing a local observation state tensor; s4, extracting an individual Q value tensor; s5, extracting a global team Q value; s6, generating an anti-fact state through the VRNN model, calculating time sequence differential loss, and jointly optimizing parameters of the whole QMIX model; and S7, based on the individual Q value tensor, selecting an evacuation guide action. The method overcomes the limitation that a traditional crowd guiding method depends on a fixed plan, the guiding strategy is single and response lags behind, and an efficient and accurate solution is provided for intelligent crowd flow management and control and cooperative guiding of large public places.

The invention discloses an electronic coupon algorithm recommendation system based on a specific crowd, which belongs to the technical field of electronic coupon recommendation and comprises a user demand perception module, a key factor priority ranking module, an electronic coupon strategy generation and scene adaptation module and an evaluation and feedback module. According to the electronic coupon algorithm recommendation system based on the specific crowd, key factors are generated and sorted by mining user preferences and purchase frequencies, and priorities are dynamically adjusted when conversion does not reach the standard; pushing associated coupons N days before exhaustion in combination with commodity consumption prediction; optimizing access based on the user active window and the channel response rate; the method solves the problems of pushing dislocation, opportunity loss and low conversion efficiency of a traditional system, realizes optimal configuration of marketing resources and improvement of user conversion efficiency, and is suitable for a precise electronic coupon marketing scene.

The present application relates to a training method, apparatus, computer equipment, computer-readable storage medium, and computer program product for a crowd detection model. The method comprises: obtaining unlabeled features and labeled features obtained from capturing a crowd scene from multiple perspectives; predicting the uncertainty of the number of people based on the unlabeled features of each group of perspectives to obtain the uncertainty of the number of people corresponding to each group of perspectives; each group of perspectives contains a different number of perspectives; comparing the uncertainty of the number of people corresponding to each group of perspectives to obtain a comparison result; determining the unlabeled loss based on the comparison result; and training a crowd prediction model based on the unlabeled loss and the labeled loss corresponding to the labeled features. This method can alleviate the dependence on labeling and improve the accuracy of crowd prediction.

The invention is suitable for the field of computer vision, provides a crowd evacuation path dynamic recommendation method and system based on visual data, and discloses a crowd evacuation path dynamic recommendation method and system based on visual data in the field of computers. Crowd motion video streams are collected in real time, pedestrian head orientation vectors are extracted after preprocessing, a crowd motion velocity field is formed, and a crowd motion state database is established; constructing a visual inertia dynamic model to predict future crowd evacuation path distribution change; the real-time bearing rate of each channel is calculated, and grading early warning is generated by combining capacity threshold pre-judging saturation time; and when the channel reaches the optimal capacity threshold value, determining an upstream branch node, calculating a diversion starting time window and strength, generating a control instruction, and transmitting the control instruction to an indication execution end. Accurate dynamic path recommendation is realized, and evacuation safety and efficiency are improved.

The invention provides a commercial complex crowd management system based on a real-time LoT sensor, and relates to the technical field of intelligent management. The management main system comprises a data acquisition module, a data processing module, a dynamic environment optimization module, a people flow guidance and excitation module, a resource scheduling module, a safety monitoring and response module, an external cooperation module, a space value optimization module, a user feedback and experience module and a control flow management module; the real-time LoT sensor is deployed to comprehensively monitor the people flow density, the environmental parameters and the equipment state in the commercial complex, the problems of crowded areas and uneven resource allocation in the peak period can be found in time, and user experience decline and potential safety hazards caused by delayed response are effectively avoided. The system achieves minute-level density monitoring in peak hours, and the real-time performance and response efficiency of crowd management are remarkably improved.

This invention belongs to the field of computer simulation of crowd evacuation, and proposes a crowd congestion prediction method and system based on a spatiotemporal fusion neural network. The method includes: acquiring scene information and crowd state information of crowd movement; constructing a knowledge graph based on the scene information and crowd state information, and preprocessing the knowledge graph; inputting the preprocessed knowledge graph into a graph neural network to obtain a spatial feature matrix; inputting the preprocessed knowledge graph, spatial feature matrix, and crowd state information into the spatiotemporal fusion neural network to predict the movement speed of crowds within a region, and predicting the crowd congestion situation within the region based on the crowd movement speed. The knowledge graph-based construction method of this invention can more comprehensively store heterogeneous information related to crowd movement. Furthermore, using a knowledge graph to represent scene information and store crowd movement state attributes provides relatively complete input information for the next step of the neural network to mine the spatiotemporal correlation of crowd movement.

Examples provide a video security system including an electronic processor configured to detect an anomaly by obtaining video data captured using an image sensor and determining a number of people detected in each of a plurality of frames of the video data. For each moment t, the electronic processor determines an average number of people n over a time interval (t-Δt1, t), where Δt1 is greater than or equal to the moment t. For each non-overlapping time interval Δt1, the electronic processor updates a histogram based on the average number of people for the time interval Δt1. In response to determining, using the histogram, that the probability ρ of the average number of people n over the time interval (t-Δt1, t) is less than a threshold probability ρ*, the electronic processor generates an alert indicating that an anomalous crowd size is detected.

The application discloses a kind of multi-modal crowd density prediction methods based on time convolution network, comprising the following steps: video information module obtains image from monitoring camera, the image obtained is carried out crowd counting by crowd counting model, obtains the crowd density of each camera every time, and is organized into time series data;The time series data of crowd density is respectively extracted into the current time hidden vector for each sub-region by time convolution network;Itinerary planning module extracts corresponding itinerary information from itinerary planning table for multi-modal prediction;The features of fusion module and video information module are fused, and the crowd density prediction value of each sub-region in the future is obtained.This application can improve the prediction accuracy of the model by fusing multi-modal information, and can give a smooth prediction result when facing low SNR data, and can also respond in time when the crowd density changes suddenly.

The method and system for generating an audience label of a target account provided in the specification obtain a target label from a plurality of known labels corresponding to all users of a target platform, the target label corresponding to a target group of users in the all users, calculate a group tendency value of the target group of users with the determined label to the target account, and then generate an audience label of the target account based on the target label. The target group of users in the specification has a determined label. From the perspective of the determined label, the group tendency value is determined, so that the target group of users has stronger explainability and larger operation space.

The application discloses a kind of data processing method, data processing system and storage medium, it is related to data processing technical field, method includes: client obtains target object selected target crowd label, obtains the test account preconfigured for label, determine target test account from it and execute automatic login;Forwarding server intercepts interface request sent to simulation server in application running;When request belongs to preset interface request set, it is forwarded to simulation server;Simulation server obtains matched preset response data from shared expected data pool and returns client;Client provides it to application, so that application presents corresponding interaction interface or executes corresponding function logic.The application automatically completes the whole process from crowd label selection to interface presentation by preconfiguring test account and shared expected data pool, avoids different personnel repeatedly constructing account and preparing data, solves the problem of low experience efficiency caused by repetitive labor, and improves experience efficiency.

The invention relates to a real-time crowd panic probability identification method and system based on multi-modal fusion. The method comprises the following steps: collecting video data and audio data and carrying out data preprocessing; inputting the preprocessed video data into the CDNet to obtain a density map; candidate panic areas in the density map are extracted; identifying a target person in the candidate panic area in the video data frame by frame to obtain a motion trail sequence; obtaining the motion trail panic probability of the target person according to the high-order time sequence characteristics of the motion trail sequence; converting the preprocessed audio data into a frame-by-frame text sequence, and identifying panic language content; fusing the spatial structure features of the density map, the feature vector of the panic probability of the motion track and the context semantic features of the panic language content into a unified representation vector; and inputting the unified representation vector into a multi-layer perceptron to obtain a crowd panic probability identification result. Compared with the prior art, the method has higher adaptability and robustness.

The application discloses a kind of dense crowd corresponding relation interpretation system and method based on point cloud matching, the method includes: dense crowd corresponding relation interpretation method based on point cloud matching, with the upper left corner of each scene image as origin, respectively with right and downward as positive direction respectively establish x axis and y axis, so that each node in pedestrian point cloud has only pixel coordinate expression;Pedestrian point cloud matching network extracts pedestrian point cloud feature, finds pedestrian point cloud corresponding relation, and the input is the feature tensor of two pedestrian point clouds;Interpretation result correction module modifies result after obtaining pedestrian node corresponding matrix.This method has practical value in analyzing the trend of multiple behavior individuals in dense crowd, preventing major public safety accidents, and is closely related to the construction of a safe city, the realization of intelligent monitoring and processing of dense crowd and public danger.

This invention belongs to the field of intelligent transportation, specifically relating to a MaaS-IoT integrated method and system for crowd risk perception and intervention. The method first collects environmental data and travel orders for the target area through IoT devices and an online platform; then, it constructs a path decision model based on the Bandit algorithm; combining shared and unique feature vectors to achieve path recommendation. Next, it constructs a crowd risk event dataset based on historical data and generates an easily searchable vector database, and fine-tunes the LLM to obtain a crowd risk inference model. Then, it tokenizes the time-series data, and uses the crowd risk inference model to perform chain reasoning and Monte Carlo-Shapley attribution to generate corresponding results; finally, it generates intervention instructions based on the outputs of the two models, and issues them for execution after verification. This invention solves the problem that existing methods struggle to achieve timely risk perception and efficient intervention in public places where crowds gather.

The invention belongs to the technical field of crowd counting, and particularly relates to a domain-oriented adaptive crowd counting energy-driven active learning method, which comprises the following steps of: extracting multi-scale visual features from an input image, and training HRNet to obtain an optimal source domain model; source domain and target domain samples are mapped to a unified energy space through an energy function; performing data enhancement on the target domain sample, and calculating uncertainty and a predicted people number mean value under various data enhancement modes; calculating sample energy of the target domain; screening the target domain samples twice by adopting an active learning strategy, and labeling the screened samples; designing energy alignment loss; and performing fine tuning on the optimal source domain model, and obtaining a final crowd counting result by using the fine-tuned model. According to the method, data distribution of the source domain and the target domain is effectively aligned through the energy model, sample labeling is carried out in combination with an active learning strategy, and the cross-domain crowd counting performance is improved.

The application discloses a crowd counting method and device, a terminal equipment and a storage medium. The method comprises the following steps: acquiring a target picture; performing prediction on the target picture by using a pre-trained crowd counting model to obtain a density map prediction value and a head number prediction value; and performing weighted calculation on the density map prediction value and the head number prediction value to obtain a crowd quantity count value. The head number prediction value is used as an auxiliary result to be weighted with the density map prediction value, so that the accuracy of the obtained crowd quantity count value is improved, and the accuracy of crowd counting is improved.

The invention discloses a resource optimization method and apparatus, a computer device and a storage medium. The method comprises the steps of obtaining a to-be-managed crowd and current crowd feature data and historical crowd feature data of the to-be-managed crowd; providing the current crowd feature data and the historical crowd feature data to a preset prediction model, so that the prediction model outputs a resource operation state index of the crowd; determining a resource scheduling priority of the to-be-managed crowd based on the resource operation state index; determining a target resource optimization strategy of the to-be-managed crowd based on a mapping relationship between the resource scheduling priority and a resource optimization strategy; and performing resource optimization processing on the to-be-managed crowd based on the target resource optimization strategy. Resource allocation such as computing power, bandwidth and task execution rhythm can be adaptively adjusted under different service loads and crowd scale changes, and coordinated optimization of multi-dimensional system resources and improvement of overall operation efficiency are achieved.

The application provides a crowd density estimation method and system based on feature-aware weighted contrast learning, relates to the technical field of crowd density estimation methods, and comprises the following steps: inputting a basic feature map into a constructed multi-level parallel hole convolution layer respectively, normalizing an importance score to obtain a fusion weight coefficient of each branch, and obtaining a final fusion feature map; constructing a regression network to generate a low-resolution density map, inputting the density map into a lightweight convolution network to generate a scaling mask, multiplying the density map after bilinear upsampling with the scaling mask element by element, and generating a predicted density map. The multi-level parallel hole convolution layer enables the network to simultaneously capture local details and wide-range context information, the channel attention automatically strengthens the feature channels related to the crowd, the spatial attention accurately locates the high-density area, the density map is calibrated to the actual number of people for each pixel through a loss function, and the number of people in the whole image can be directly obtained by summation.

The invention discloses a distribution method and system based on crowd density monitoring in disaster emergency hedging transfer. The method comprises the following steps: acquiring personnel position and moving speed data in real time through a sensor network deployed at a hedging channel and a temporary gathering point; identifying and quantifying a personnel gathering area, and positioning an excessive gathering area; the boundary, the people flow direction and the number of people of the excessive gathering area are extracted, and a plurality of alternative shunting paths leading to the standby gathering point are generated in combination with the topological structure of the peripheral channel, the available width and the obstacle information; simulating a shunting process to predict the future load of each path and the gathering point, and selecting a final shunting path from the alternative paths according to a load balancing principle to form a shunting scheme; and finally, integrating into a global transfer path diagram, generating an updated evacuation path, and executing guidance. According to the invention, real-time sensing, dynamic path planning and global load optimization of the crowd gathering risk are realized, so that the evacuation efficiency and the overall safety are improved.

The embodiment of the invention relates to a multi-center badminton action evaluation method based on a trusted execution environment, which comprises the following steps: acquiring an action fragment to be evaluated, and locally extracting an action feature vector associated with the action fragment to be evaluated to obtain an action feature to be evaluated; performing cooperative retrieval with at least one other participation center based on the to-be-evaluated action features to obtain a candidate action tag set; and based on the candidate action label set and a preset action evaluation model, obtaining an evaluation result of the to-be-evaluated action fragment. According to the technical scheme provided by the embodiment of the invention, on the premise of not converging original video data and not converging each center standard / auxiliary database, a privacy calculation and security aggregation mechanism is introduced, cross-center collaborative retrieval and evaluation are realized, and federated iterative optimization of a global model is supported. Therefore, the privacy of the trainer and the security of database assets can be guaranteed, and normative evaluation and deviation positioning adaptive to different crowds and training stages can be provided at the same time.

The invention discloses a multi-crowd-dependent task scheduling method and device, and relates to the technical field of portrait crowd processing, whether a crowd calculation task meets an execution condition is judged according to a dependency label detection result and a dependency label detection result, and if the execution condition is met, the crowd calculation task is scheduled. The crowd calculation tasks placed in the waiting queue are traversed according to the specified interval through the dependency detection thread so as to perform dependency detection, so that the dependency crowd detection result and the dependency label detection result are ensured to be in a ready state during crowd calculation, and the crowd calculation efficiency is improved. The influence on the accuracy and freshness of crowd calculation due to problems such as missing dependent crowd detection and dependent label detection is avoided, and if the crowd calculation task meets an execution condition determined by a dependent crowd detection result and a dependent label detection result, the crowd calculation task meeting the execution condition can be put into a task queue to be executed in time; and data depending on monitoring is stable and reliable, so that the quality and accuracy of the data are improved in the process of solving the dependency problem of portrait crowds.

The invention discloses a dense crowd counting method based on an improved yolov11 model. The dense crowd counting method comprises the following steps: improving a feature extraction network of a yolov11n model; replacing an original CBS and C3K2 module of the yolov11n model with an encoder module of a U-NET V2 network; the feature fusion network of the yolov11n model is improved; an AIFI module is added behind the highest layer of the feature fusion network; a detection head of the yolov11n model is improved, and a DIOU loss function is adopted to replace an original loss function; and inputting the preprocessed image tensor into the improved yolov11n model for training, and storing the trained model for dense target counting detection. According to the method, the calculation speed is ensured as much as possible, and the precision of crowd number detection when severe shielding exists between targets in a dense scene is improved.