180results about How to "Effective classification" patented technology

The invention belongs to the technical field of data security, and discloses a data management method based on a data resource security identification level, and the method achieves the automatic and intelligent decision making: achieves the data classification and the automatic identification of the security level through a machine learning and deep learning model, remarkably reduces the manual intervention, and improves the efficiency and accuracy. And dynamic real-time performance: in combination with a risk assessment model, the system can detect abnormal behaviors in real time and dynamically adjust a security policy in a data flow process, so as to ensure that the sensitive data is properly protected in different scenes. And security and compliance: a log auditing module ensures the traceability of all data operations, improves the response capability to potential risks through an anomaly detection algorithm, and provides effective compliance guarantee for data security governance. And the system can be compatible with various data sources, supports configuration and execution of various security policies, and is suitable for data security management requirements of different industries (such as finance and medical treatment).

The invention discloses a multi-modal classification method fusing a graph convolutional neural network and a capsule graph neural network, and solves the problem of how to apply an artificial intelligence technology to an appearance patent. The method comprises the following steps: coding a patent text sequence by utilizing BERT, segmenting semantic graph convolution by utilizing PCNN, constructing a dependency tree by utilizing grammar graph convolution, carrying out interactive attention and relationship classification, and finally carrying out image feature extraction through a capsule network. According to the method, the appearance patent classification technology fusing heterogeneous multi-modal information is researched, the feature extraction and fusion problems of multi-modal information such as characters, images and networks are solved, a heterogeneous multi-modal appearance design patent classification model is constructed, and the classification accuracy is greatly improved relative to pure text content. And meanwhile, an appearance design patent AI standard data set is constructed on the basis of the intellectual property hypergraph network, so that the blank of patent types in a traditional Lankyo classification system is filled, and the bottleneck of a traditional appearance patent classification method is broken through.

The invention provides a network node classification method based on a multilayer graph neural network model, and the method comprises the steps: constructing a graph according to network nodes, carrying out the preprocessing of initial node features in the graph, and obtaining an initial embedded vector; according to the initial embedding vectors of the node and the neighbor node thereof, updating the node representation by adopting a layered message passing mechanism; according to the initial embedded vectors of the nodes and the neighbor nodes thereof, the graph attention network uses an attention mechanism to calculate weights for the neighbor nodes of each node and then dynamically updates feature vectors of the nodes; a multi-head attention mechanism is adopted to splice and merge the feature vectors of the nodes output by the graph attention network and the updated node representation, and the comprehensive feature vector of each node is obtained; the comprehensive feature vectors are further optimized through graph contrast learning and a node embedding module, and optimized features are obtained; and obtaining predicted node classification according to the optimization features. According to the method, complex features of heterogeneous graph data can be effectively captured, the method has good robustness for an over-smoothing problem, more accurate heterogeneous network representation learning is realized, and the method has advanced performance and robustness in a node classification task.

The invention provides an LED spectrum fingerprint open set identification method based on Transform and support vector data description and related equipment. The method comprises the following steps: collecting an optical signal emitted by at least one LED device, converting the optical signal into spectral data, and dividing the spectral data into a training set and a test set; inputting the training set into a preset feature extraction network to carry out network training so as to learn and obtain a spectral feature vector; wherein the feature extraction network comprises a Transform branch, a CNN (Convolutional Neural Network) branch and a feature interaction module; based on the spectral feature vectors of the training set, constructing a hypersphere model by using an SVDD algorithm, and optimizing parameters of the hypersphere model to obtain an optimal hypersphere model; and inputting test samples in a test set into the trained feature extraction network to obtain corresponding spectral feature vectors, and inputting the spectral feature vectors of the test samples into the optimal hypersphere model for classification.

The invention discloses an abnormal defect detection method and system applied to PCB intelligent welding, and relates to the technical field of PCB welding detection, and the method comprises the following steps: carrying out the image collection of a whole PCB through a high-definition camera, and recognizing and positioning all welding spots on the PCB through an image segmentation technology; and each welding spot is detected according to the initially set expected resolution, and after the welding spots with obvious welding problems are recognized, the recognized welding spots are screened out. According to the method, the welding spots are accurately positioned through the image segmentation technology, the detection efficiency is rapidly screened and improved by using the initial resolution, and obvious non-pseudo-soldering welding spots and potential pseudo-soldering welding spots are efficiently classified in combination with a deep learning model. Aiming at potential cold solder joints, the resolution is dynamically adjusted, high-resolution fine detection is performed, the cold solder problem is accurately identified, the detection accuracy and precision are remarkably improved, the omission ratio and the misjudgment ratio are reduced, and reliable technical support is provided for PCB intelligent welding quality control.

The invention discloses a pet peculiar smell detection method and device and a storage medium. The pet peculiar smell detection method comprises the following steps: collecting data in a to-be-detected environment in real time through a sensor array; performing temperature compensation on the sensing signal through a weighted linear temperature compensation algorithm and a weighted adaptive whale optimization algorithm; setting a threshold value, comparing the compensated sensor signal amplitude with the set threshold value, and when any sensor signal amplitude is triggered, starting to collect sensor data with a fixed length; inputting the collected sensor data with the fixed length into a self-adaptive time convolution GSA-TCN network model based on a gating convolution kernel rarefaction strategy for classification and identification; when it is judged that the pet excretes peculiar smell, table look-up, interpolation and concentration calculation are carried out according to the maximum value of the sensor data with the fixed length, and the detection result and the concentration value of the environment to be detected are output. The pet peculiar smell can be effectively detected, a series of problems caused by the pet peculiar smell can be solved, and the application scene is wide.

The invention belongs to the technical field of fault diagnosis of wind power generation equipment, and relates to a fan gearbox fault image recognition method and system. According to the method disclosed by the invention, the multi-modal data of the fan gear box is acquired, and the multi-modal data complement each other, so that the accuracy and reliability of the data on fault characterization are improved. And carrying out deep learning analysis on the preprocessed multi-modal data to obtain a deep learning result, and carrying out efficient and accurate classification on fault types of the fan gearbox by training a deep learning model. Fault reasoning is carried out based on expert knowledge and historical data to obtain a knowledge graph reasoning result, the expert knowledge and the historical data are combined, a knowledge graph is constructed, knowledge fusion and sharing are realized, and logical reasoning is carried out by using the knowledge graph. Fault diagnosis is carried out according to a deep learning result and an identification map reasoning result, the advantages of various methods are integrated, the fault type and part can be identified more quickly, and timely guidance is provided for subsequent maintenance work.

The invention relates to a visual analysis method and device for multiple internal corrosion failure types of a pipeline. The method comprises the following steps: inputting internal corrosion failure picture data to be recognized into an internal corrosion failure type recognition model to obtain a predicted internal corrosion failure type and a feature map of the last convolutional layer; analyzing the predicted internal corrosion failure type and the feature map of the last convolutional layer by using Grad-CAM to obtain a feature region thermodynamic diagram; based on the predicted internal corrosion failure type and the characteristic region thermodynamic diagram, obtaining the contribution degree of the characteristic of each characteristic region to an identification result; wherein the internal corrosion failure type identification model is obtained by the following steps: obtaining a pipeline internal corrosion failure data set; inputting the internal corrosion failure data set into an initial internal corrosion failure type identification model for training to obtain a trained internal corrosion failure type identification model; wherein the initial internal corrosion failure type identification model is constructed by using a Resnet18 convolutional neural network.

The invention provides a process optimization method for coloring an artificial fur fabric by using cations, which belongs to the technical field of artificial fur fabric coloring processes, and comprises the following steps: collecting various process parameters and quality index data required by artificial fur fabric coloring, including grey cloth specification, squeezing treatment, dye formula, dyeing process, post-treatment and the like. Establishing a coloring effect empirical model considering each process parameter, wherein the coloring effect empirical model comprises equations of dye diffusion, fixation, fiber water absorption, fluff shaping and the like; substituting the collected parameter data into the model to obtain a preliminary process parameter-effect relationship; a large number of technological parameter combinations are generated according to the parameter range, and clustering analysis is carried out by utilizing a model calculation theoretical effect. And selecting process parameters of the typical clustering center to carry out pilot test verification, and optimizing the process parameter-effect relation model. And adopting a genetic algorithm to obtain an optimal process parameter combination on the basis of the optimization model, carrying out pilot scale production verification, and finally determining an optimization process scheme.

The embodiment of the invention discloses a dangerous chemical identification method, and the method comprises the steps: obtaining chemical information data, and constructing a dangerous chemical semantic database according to the chemical information data through employing a deep learning multi-modal data fusion algorithm; according to naming rules and attributes of chemicals, defining text variables and numerical variables, extracting hazardous chemical judgment conditions from the chemical information data, and establishing a hazardous chemical rule base according to the text variables, the numerical variables and the hazardous chemical judgment conditions; according to the chemical information data, establishing a hazardous chemical data model based on a graph neural network technology; constructing a hazardous chemical substance identification system by using the hazardous chemical substance semantic library, the hazardous chemical substance rule library and the hazardous chemical substance data model; and obtaining to-be-classified chemical information, and classifying to-be-classified chemicals in the to-be-classified chemical information by using a hazardous chemical identification system to obtain a hazardous chemical judgment result. According to the invention, more intelligent and accurate dangerous chemical identification is realized.

The invention relates to a data mining method, a multi-modal data query method and a vertical domain knowledge question and answer method based on a knowledge graph. The mining method comprises the steps of generating a knowledge graph based on a first artificial intelligence model; inputting initial node information acquired by the knowledge graph into a second artificial intelligence model, wherein the second artificial intelligence model comprises a knowledge graph expansion unit, an abnormal node acquisition unit, a key node acquisition unit and a value mining unit; expanding expanded node information from the initial node information through a knowledge graph expanding unit, and respectively outputting the expanded node information to the abnormal node acquisition unit and the key node acquisition unit; and the abnormal node acquisition unit and the key node acquisition unit respectively output abnormal nodes and key nodes in the knowledge graph to the value mining unit, so that the value mining unit presents related information of the abnormal nodes and the key nodes.

The invention relates to the field of biological recognition and environmental monitoring, and discloses a chironomidae larva recognition system based on deep learning and image recognition, which comprises an image acquisition module, an automatic sample processing device, an image superposition processing module, a deep learning recognition module and an automatic category classification and statistics module, the image acquisition module acquires multi-focal plane chironomidae larva images through an optical microscope and a digital camera, the automatic sample processing device realizes automatic loading operation of a sample, and the image superposition processing module performs adaptive fusion on the multi-focal plane images to generate a full-focal plane image containing multiple pieces of depth information; the deep learning recognition module is used for extracting fine structure features of chironomidae larvae and automatically recognizing and classifying different types of chironomidae larvae, and the automatic type classification and statistics module is used for performing statistical analysis according to classification results and generating a report of type distribution and time sequence change. The method is suitable for automatic monitoring of water ecological environment.

The invention discloses a big data real-time processing system and method, and relates to the technical field of big data processing, and the big data real-time processing system comprises a data partitioning module, a data storage module, a data reading module and a data backup module; the data partitioning module is used for partitioning big data entering a block chain in real time by using a partitioning strategy, extracting keywords, and distinguishing the frequency of data blocks according to historical query records; the data storage module stores data blocks through a storage contract, encrypts identifiers of the data blocks and establishes a mapping relation; the data reading module screens data blocks based on the private key set when a user queries, and determines a sending sequence through keyword matching; the data backup module generates a backup after storage, performs different updating strategies for data blocks with different frequencies, and performs adaptive modification on original data by using a hash value; in combination with the block chain technology, multiple advantages of real-time processing, efficient storage, safe reading, flexible backup and the like are integrated, and a comprehensive and efficient solution is provided for big data processing.

The invention belongs to the field of non-intrusive load monitoring, and discloses a household electricity consumption analysis method based on non-intrusive load monitoring and related equipment, firstly, non-intrusive monitoring is realized by using household total electricity meter data instead of a single equipment electricity meter, and the complexity and cost of household electricity consumption data collection are reduced; secondly, an unsupervised learning method is adopted to cluster the power load state data, data does not need to be marked in advance, and different load events are automatically identified and classified, so that the problems of load type diversity and high identification difficulty are effectively solved, and the accuracy and efficiency of load event detection are improved; in combination with a depth-first search algorithm and load event constraint conditions, fine search is carried out on load event labels, an undetermined event sequence set is generated, the logicality and continuity of event sequences are enhanced, electricity utilization behavior analysis of the electric equipment is deeper and finer, home users can better understand and optimize own electricity utilization modes, and the user experience is improved. And energy conservation, emission reduction and efficient utilization of energy are promoted.

The invention provides a method and a device for evaluating the grade of a butter crab based on spectral information, and relates to the technical field of engineering optics. The method comprises the following steps of: performing spectral measurement on the butter crab to be evaluated by using a spectrograph to obtain spectral data, extracting characteristic parameters from the data, and then calculating the similarity to obtain the grade of the butter crab; and selecting the quality grade corresponding to the maximum similarity value as the initial quality grade of the butter crabs to be evaluated, wherein the grades comprise top-grade products, special-grade products, first-grade products and unqualified products. And further combining the characteristic parameters, calculating the yellow-blue value and the light transmission ratio area of the to-be-evaluated butter crab, and meanwhile, considering the fatness and ovary index to form a comprehensive evaluation index. And finally, comparing the evaluation index with a preset quality interval, correcting the to-be-determined quality grade according to a comparison result, and finally determining the actual quality grade of the to-be-evaluated butter crab. According to the method, comprehensive evaluation of the quality of the butter crabs is realized through a spectral measurement technology and a characteristic parameter extraction method.

The invention discloses a basin-type insulator internal defect detection system and method based on pulse infrared thermal waves, and relates to the technical field of nondestructive testing and computer vision, and the basin-type insulator internal defect detection system comprises a pulse infrared thermal wave excitation module, an infrared thermal imaging module and a data processing module; the pulse infrared thermal wave excitation module is used for enabling the surface of the insulator to generate thermal response in a pulse infrared thermal wave excitation mode; the infrared thermal imaging module captures the temperature change of the surface of the insulator in real time and performs early detection; and the data processing module is used for data preprocessing, feature extraction and defect detection. According to the method for detecting the internal defect of the basin-type insulator based on the pulse infrared thermal wave, the pulse infrared thermal wave excitation and the high-precision infrared thermal imaging technology are combined, so that the bubble defect in the insulator is efficiently detected. The pulse infrared thermal wave technology generates instantaneous temperature change on the surface of the insulator, so that the bubble defect is more obvious, and the detection sensitivity and speed are improved.

The invention discloses a model pruning method based on channel attention and adaptive learning, and the method comprises the steps: firstly introducing a channel attention mechanism SENet, adding the SENet to each convolution layer of a target model, and training the SENet until the model converges; then, extracting the middle output of the SENet of the training sample, combining with a scaling factor in a BN layer to obtain a channel importance score, and then performing pruning according to the channel importance score; and finally, training the pruned model by using an adaptive learning loss function until convergence. According to the method, the model pruning principle and the channel attention principle are fully utilized, and the high-precision model pruning method in the remote sensing field is achieved. The method can greatly reduce the consumption of computing resources, improve the processing efficiency and reduce the hardware cost while maintaining the model precision or even exceeding the model precision before pruning.

The invention discloses an ecological environment protection system for sorting management of historical remaining solid wastes, and relates to the technical field of environment protection, and the ecological environment protection system comprises a solid waste collection module, an intelligent sorting module, a treatment and disposal module, an environment monitoring module and a central control module; the solid waste collecting module is used for collecting solid waste which is left in history and comprises different types of garbage storage containers, sealing and transporting the solid waste, and transporting the collected solid waste to a sorting treatment area. According to the ecological environment protection system, a plurality of modules are integrated, an efficient and scientific historical remaining solid waste sorting management system is constructed, the system can accurately predict the waste output, the design of a waste storage container is optimized, effective collection and sealed transportation of waste are ensured, and the system is suitable for large-scale popularization and application. And the intelligent sorting module is combined with an image recognition technology and a sensor technology, so that the sorting accuracy is improved, and resource utilization of wastes is facilitated.

The invention belongs to the technical field of natural language processing, and discloses a natural language processing platform and method based on deep learning, and the platform comprises a data collection module which is used for collecting font data, text data, image data and environment data; the data processing module is used for preprocessing the collected font data, text data, image data and environment data and obtaining a comprehensive feature data set; the character recognition module is used for predicting character recognition results through a character recognition model and judging whether the character recognition results are consistent or not; the semantic recognition module is used for generating corresponding character paraphrases through a semantic recognition model; collecting character recognition results and corresponding character paraphrases through the software terminal; the voice output module is used for converting the character recognition result and the paraphrases of the corresponding characters into natural voices and outputting the natural voices by adopting a TTS technology; through the implementation of the technology, students can be helped to better learn and understand.

The invention belongs to the technical field of artificial intelligence, and relates to a junk call identification method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining initial audio data corresponding to a target call when the target call is accessed; preprocessing the initial audio data to obtain specified audio data; performing feature extraction on the specified audio data based on a pre-training language model to obtain a specified audio feature vector; calling a junk call recognition model; performing identification processing on the specified audio feature vector based on a junk call identification model to obtain a probability that the target call belongs to each preset junk call category; and screening out a target probability with the highest value from all probabilities, and generating a junk call recognition result corresponding to the target call based on the target probability. In addition, the invention also relates to a block chain technology, and the junk call recognition result can be stored in a block chain. Based on the use of the junk call recognition model, the junk call recognition efficiency and recognition accuracy are improved.

The invention discloses a network intrusion detection method based on clustering oversampling and Transform, and relates to the technical field of network security and deep learning. The method comprises the following steps: carrying out data cleaning on network flow data, balancing a data set through a clustering oversampling technology, and extracting characteristic values through a method of combining information gain and rapid correlation filtering; normalizing the feature values by adopting a Max-Min normalization method, converting the feature values into images according to timestamps and feature sizes, and marking attack types; amplifying the generated image by adopting a bilinear interpolation method, and dividing the image into a training set and a test set; and fine tuning is carried out on the classification head through transfer learning, deep features of network traffic data are extracted, and accurate classification of normal traffic and various attack traffic is realized. According to the network intrusion detection method based on clustering oversampling and Transform, different types of network attacks can be effectively identified, and the accuracy rate and reliability of intrusion detection are remarkably improved.

The invention relates to the technical field of music information retrieval, in particular to an automatic classification method and system for music works, and the method comprises the following steps: based on the classification standard of music genre, analyzing the hierarchical node relationship of the music genre, extracting the corresponding hierarchical characteristics of the large class and the small class of the genre, identifying the association proportion between each hierarchy, and analyzing the hierarchical connection strength. And generating a music genre level weight table. According to the method, through deep analysis of a music genre hierarchical structure, finer and dynamic classification of music works is realized, the classification accuracy and flexibility are significantly improved, through extraction of hierarchical features and revelation of association between hierarchies, music classification not only breaks through limitation of simple labeling, but also can reflect hierarchical relationships between genre, and the classification efficiency is improved. The chord change rule and amplitude are extracted, the dynamic capture of music features is enhanced, the association of chord change and genre level features is matched, the classification weight is dynamically adjusted, genre features are accurately matched, and the diversity and complexity of music creation are better adapted.

The invention belongs to the technical field of artificial intelligence, and discloses an FPGA fault diagnosis accelerator design method based on a shift module and an AXI cross matrix, and the method comprises the steps: taking GHOSTNETV2 as a backbone network, carrying out the standard 3 * 3 two-dimensional convolution operation of an input content through a first main convolution layer of the network, so as to extract features, and carrying out the calculation of the features; then inputting the features into a subsequent layer for further feature extraction, and independently optimizing the layer; the influence of quantization and data packaging on delay is analyzed, and different quantization methods of different data formats from 32 bits to 8 bits are compared; different quantization strategies are selected for different network layers; an advanced synthesis tool HLS of the Xilinx FPGA is selected; each Xilinx DSP48E1 is decomposed into two 8 * 8 multiplying units, so that computing resources are maximized. According to the method, an optimal mixing precision quantification strategy is obtained through theoretical analysis and experiments, and knowledge distillation is decided to be used for keeping precision, so that the use of hardware resources is remarkably reduced.

The invention discloses a data storage method and system for a storage chip, and the method comprises the steps: selecting reference data and sample data from target storage data, and carrying out the analysis, and obtaining a data similarity value; the data similarity value is compared with a data similarity threshold value, and if the data similarity value is larger than or equal to the data similarity threshold value, the reference data and the sample data are classified into the same classification data set; obtaining a residual storage space, judging whether the residual storage space is smaller than a residual storage space threshold value or not, and if yes, analyzing data in the classified data set to obtain a data set key value; identifying the non-key data group to generate a compression optimization signal; based on the compression optimization signal, judging whether the residual storage space of the system is smaller than a residual storage space threshold after compression processing, and if yes, analyzing the non-high access frequency data to obtain a compression priority value; and performing optimization compression on the non-high access frequency data according to the compression priority value.

A fault diagnosis and state evaluation method for GIS ultrahigh frequency partial discharge signals is characterized by comprising the following steps: S1, based on a physical model of GIS partial discharge typical defects, adopting a specific test loop to carry out a discharge test, collecting ultrahigh frequency partial discharge signals of multiple power frequency periods of different defects, grouping the signals, and carrying out fault diagnosis and state evaluation on the signals; each group comprises a predetermined number of power frequency periodic signals; s2, calculating a partial discharge statistical characteristic quantity, and performing normalization processing to form a data sample corresponding to each group of partial discharge signals; s3, the data samples are regarded as observation values in clustering, different total clustering numbers k are selected, K-means clustering analysis is carried out on each k value, and in-cluster distance sum of squares (WCSS) under each clustering result is calculated; S4, a relation curve of WCSS-k is drawn, and an elbow rule is adopted to preliminarily judge the optimal clustering number; and S5, calculating absolute values of Pearson's correlation coefficients between the WCSS and k under different k values, and when rho begins to rise to be close to 1, determining an optimal clustering number kopt as final clustering output. The optimal clustering number kopt determined by the method can effectively distinguish and classify GIS partial discharge ultrahigh frequency signals, so that different types of discharge signals are distinguished, and powerful support is provided for effective diagnosis of GIS partial discharge defects.

The invention provides an algae target detection method, system and device based on deep learning, and relates to the field of computer vision environment monitoring, and the method mainly comprises the steps: collecting the image data of a water body through an image sensing device; splitting and preprocessing the image data according to a preset size; performing small target splitting and data enhancement on the training image data; based on the training image data, training the optimized YOLO target detection model; and performing target detection on the detected image data through the trained YOLO target detection model to obtain a detection result. According to the scheme, by introducing and optimizing a YOLO series algorithm, the automation degree and the detection precision of algae detection can be remarkably improved; according to the scheme, algae target detection and classification can be accurately and efficiently carried out in complex water backgrounds and various algae environments, and the false detection rate can be effectively reduced; the scheme is suitable for real-time monitoring and analysis scenes of a large-scale and complex water body environment.

The invention discloses an intelligent garbage classifying, compressing and recycling device, and relates to the technical field of garbage classifying and recycling devices.The intelligent garbage classifying, compressing and recycling device comprises a shell box body, a garbage feeding opening is formed in the top of the shell box body, and a bag breaking mechanism is arranged under the garbage feeding opening; a screening groove connected with an oscillating mechanism is arranged under the bag breaking mechanism, a shifting wheel is arranged in the screening groove, a master control system is arranged below the screening groove, an ejection type conveying belt is arranged on the left side of the screening groove, and a fan, a vacuum suction cup set connected with a movable telescopic mechanism and the telescopic mechanism are arranged above the ejection type conveying belt. A compression mechanism is arranged below the ejection type conveying belt; according to the device, garbage bag breaking, conveying, classifying and compressing can be fully automatically completed, the preliminary classifying effect is remarkable, the functions are comprehensive, and the universality is high; the mechanical structure is simple and reasonable, and the popularization value is high.

The invention is suitable for the technical field of sleep apnea detection, and provides a sleep apnea detection method based on wavelet transform. According to the method, the discrete wavelet decomposition technology is adopted, electrocardiogram signals are decomposed into multi-scale information to serve as model input, the supervised comparative learning concept is fused, an end-to-end deep learning network model is constructed, and efficient and accurate classification of sleep apnea is achieved. Compared with a traditional detection method which depends on an expert manual analysis method or needs complex preprocessing steps, the detection efficiency is improved, and the complexity of signal processing is simplified. More importantly, the time sequence signals are directly used as input, the complex secondary feature extraction process is avoided, and therefore the method can be converted into actual hardware equipment more easily to be implemented and applied. In addition, a disclosed Apnea ECG data set is used for testing, and the prediction accuracy of the method exceeds that of a current automatic sleep apnea detection model based on an electrocardiogram.

The invention discloses an intelligent household garbage classifier and a working method thereof, and belongs to the technical field of garbage classification. The intelligent household garbage classifier comprises an external structure, an internal structure and a control system, the external structure comprises a shell and a pulley; the internal structure comprises an intelligent identification unit, a classification execution unit, a user interaction unit, a data management unit and an environmental protection design unit; the control system comprises a visual processing system, a shunting system, a voice interaction system, a mobile application control system and an energy recovery system. The garbage classifier realizes automatic identification and classification of garbage types, and improves the classification accuracy and efficiency; the user interactivity is enhanced; the system integration is optimized, the cost is reduced, and environmental protection and sustainability are emphasized.

The invention provides an expected function safety quantification method for a multi-task automatic driving perception model, and the method comprises the following steps: 1, building a probability multi-task detection model based on MC-Dropout, and combining an MC-Dropout method with a YOLOP model; the first part of the improved probabilistic multitask detection model is an encoder network in which a CSPDarne backbone network is combined with a Dropout layer, and the second part of the improved probabilistic multitask detection model is a decoder network in which a multitask detection head is combined with the Dropout layer; 2, sampling results of the probabilistic multi-task detection model are effectively sorted and analyzed through a post-processing method, and uncertainty evaluation indexes are obtained; and 3, determining the weight coefficient of the evaluation index obtained in the above steps by adopting a mode of combining an entropy evaluation method and a layer relation analysis method. By applying the method, the scene uncertainty of the model can be comprehensively evaluated, and the expected function safety is enhanced.