20results about How to "Reduce recognition errors" patented technology

The invention discloses a policy interaction relation identification method and device, and the method comprises the steps: inputting a to-be-analyzed policy text into an element extraction model, and obtaining an element set in the policy text; if the number of the elements in the element set is greater than or equal to two, generating an element pair candidate set based on the element set; if the element pair candidate set is not empty, obtaining a relation type and a credibility score corresponding to each element pair through a relation discrimination model; comparing the credibility score with a preset credibility threshold value, and screening out a target element pair from the element pair candidate set; performing time sequence consistency verification on the screened target element pairs; and outputting the target element pair passing the time sequence consistency verification and the corresponding relationship type thereof as the policy interaction relationship of the policy text. According to the method, automatic identification of the policy interaction relationship is realized, and the accuracy and reliability of an identification result are effectively improved.

This application discloses a training method, application method, apparatus, and device for a text correction model. The related embodiments can be applied to various scenarios such as artificial intelligence to improve the accuracy of typo recognition. The method of this application includes: outputting a first text probability distribution corresponding to each character in the training sample to be corrected through a text correction model; calculating a correction loss value corresponding to the training sample to be corrected based on the first text probability distribution corresponding to each character; replacing any original character in the training sample to be corrected except for the original typo to construct a noise sample corresponding to the training sample to be corrected; obtaining a second text probability distribution corresponding to each character in the noise sample; calculating a noise loss value corresponding to the training sample to be corrected based on the first and second text probability distributions; and adjusting the parameters of the text correction model based on the correction loss value and the noise loss value to obtain the target text correction model.

This application discloses a method for training a recognition model, a recognition method, a computer device, and a computer-readable storage medium. The training method includes the following steps: controlling a single-photon lidar to emit photon pulses towards a target scene and receiving one-dimensional single-photon echoes reflected from the target scene, where the target scene includes at least one object to be recognized; constructing a training sample set based on the one-dimensional single-photon echoes; constructing an initial model and inputting the training sample set into the initial model for iterative training; validating the initial model after each iteration, marking the validated initial model as the recognition model, and outputting it. The application method involves inputting the echo signal to be recognized into the trained recognition model, and obtaining the recognition result based on the processing of the recognition model. The recognition result includes the pose and type of the object in the target scene. Therefore, this application can achieve high-precision recognition of long-distance targets.

The embodiment of the invention provides a voice recognition method, electronic equipment, a storage medium and a program product, and relates to the technical field of artificial intelligence, and the method comprises the steps: collecting a voice signal, and determining a target field related to the voice signal; obtaining a target speech recognition model corresponding to the target field, and inputting the speech signal into the target speech recognition model to obtain an initial speech recognition result, the initial speech recognition result comprising at least one character string; for each character string, under the condition that abnormal characters exist in the character string, candidate standard expressions corresponding to the character string are searched in a standard corpus, the character string is updated through the candidate standard expressions, and the candidate standard expressions are standard expressions with the highest similarity with the character string in the standard corpus; and determining a target speech recognition result according to all the character strings. According to the invention, the normalization and accuracy of a speech recognition result can be improved.

This invention discloses a method and system for identifying abnormal driver behavior based on continuous learning, relating to the field of driving behavior detection technology. The method includes dividing a driver monitoring image into image blocks and projecting them into feature vectors, adding positional encoding, and combining extracted prompt features with segmentation into prompt vectors to achieve dual coverage of global and key local features, improving feature extraction accuracy. The two types of vectors are concatenated into a fused feature sequence, input to a pre-trained Vision Transformer model with frozen parameters, and enhanced features are extracted through a self-attention mechanism. Then, a global feature vector is obtained through average pooling. Continuous learning is achieved by outputting a vector of all seen task logic values, using a binary mask for filtering, and classifying using the argmax function. This allows for adaptation to behavior category expansion, balancing real-time performance with stability in complex environments, improving the accuracy of similar behavior recognition, and making it easily deployable in in-vehicle scenarios.

This invention relates to a multimodal orbital classification and recognition method, terminal, and medium based on edge AI. The method, applied to a mobile terminal, includes the following steps: acquiring a two-dimensional color image of a face to be recognized; performing region segmentation on the two-dimensional color image of the face to obtain a color image of the orbital region; acquiring three-dimensional face data corresponding to the two-dimensional color image of the face; obtaining a face depth image based on the three-dimensional face data; performing region segmentation on the face depth image to obtain a depth image of the orbital region; and using the color image of the orbital region and the depth image of the orbital region as input to a lightweight multimodal recognition model to obtain the classification result. Compared with existing technologies, this invention has advantages such as high reliability and small scale.

The application relates to the technical field of household appliances, and discloses a cleaning brush and a control method thereof. The cleaning brush comprises a main machine, a handle and a machine head. The machine head is connected with the main machine and comprises a water tank and brush hairs. Cleaning medium is arranged in the water tank, and the brush hairs are arranged at the bottom of the water tank. An atomizing water spraying device is arranged at the bottom of the water tank and communicates with the inside of the water tank. The other end of the atomizing water spraying device is provided with a nozzle. The atomizing water spraying device is used for atomizing the cleaning medium and spraying the atomized cleaning medium to a surface to be cleaned and the brush hairs. Compared with the related art, the water tank and the atomizing water spraying device are integrated in the brush head, so that an additional handheld water pipe, a shower for humidification or frequent wetting of the brush hairs is not needed, the cleaning operation process is simplified, and the use convenience is remarkably improved.

The application provides a kind of based on image semantics increase or decrease hook problem dynamic identification and hierarchical processing method, the method includes that the increase or decrease hook project area multi-source data is unified, time alignment and grid encoding are carried out to coordinate, remote sensing image is cut according to plot boundary, internal, boundary neighborhood semantic feature and shape feature are extracted, boundary reliability is calculated and plot semantic representation is constructed, direction texture and hierarchical distribution features are extracted through double passage model, are fused into structural semantic features according to boundary reliability weighting, combine topographic elevation and historical data to extract multi-temporal time series change characteristics, generate space-time semantic vector, determine construction occupation, reclamation deficiency, boundary mismatch and use deviation problem, and update boundary reliability, determine plot problem level according to the determination result, analyze adjacent similar plot group, determine boundary adjustment scheme, output processing results containing problem level and boundary optimization result.

The present application provides a kind of based on multi-source vision sensor fusion's blast furnace slag iron flow detection method and system, belongs to blast furnace slag iron flow detection technical field, including: hyperspectral image data input slag iron ratio estimation model, the composition of molten iron and slag is identified and separated, extract the spatial distribution information of iron phase and slag phase, calculate slag iron ratio;Slag iron flow time series data input time series visual flow velocity perception model, the dynamic analysis of slag iron flow flow process, extract time evolution characteristics and flow trajectory information, calculate slag iron flow velocity;Three-dimensional point cloud data input three-dimensional point cloud reconstruction flow section analysis model, the spatial form of slag iron flow is modeled and analyzed, extract three-dimensional geometric structure and cross-sectional shape characteristics, calculate slag iron flow cross-sectional area.Slag iron ratio, slag iron flow velocity, slag iron flow cross-sectional area are jointly modeled, the adaptive calculation of slag iron flow mass flow is carried out, and slag iron flow detection result is output.The present application can effectively carry out slag iron flow detection.

The application provides a workpiece free edge automatic machining method and device, the method comprises the following steps: scanning a workpiece to be machined through a scanning module to obtain scanning data, extracting workpiece contour data of the workpiece to be machined from the scanning data, identifying a workpiece free edge to be machined in the workpiece contour data, and obtaining a machining data range of the workpiece free edge; before machining the workpiece to be machined, detecting actual position data of the workpiece free edge in the machining data range through a detection module, updating the machining data range based on the actual position data; and controlling a machining module to machine the workpiece free edge based on the updated machining data range. Through the design of automation, precision and intelligence, the application solves the problems of inaccurate identification and low efficiency in large workpiece machining.

This application provides a method and system for intelligent identification of minerals in weak information areas based on transfer learning, relating to the field of transfer learning technology. First, a source domain mineral sample set and a set of mineral samples to be identified in weak information areas are obtained. Then, geological correlation elements between the two are extracted and a cross-domain feature mapping is constructed. The original features of the minerals in the weak information areas are subjected to transfer adaptation processing to generate intermediate features of the minerals in the weak information areas that are adapted to the feature dimensions of the source domain. Next, sparse feature dimensions in the intermediate features are identified, and the corresponding feature distribution patterns of the source domain are used to complete the features, generating completed features. The completed features are input into a pre-trained transfer learning identification model for cross-domain identification, generating preliminary identification results. Finally, source domain samples with similar geological correlation elements are selected to determine the transfer deviation, correct the preliminary identification results, and generate the final identification results, effectively improving the accuracy and reliability of mineral identification in weak information areas.

This invention discloses a method and system for identifying waste materials based on multimodal fusion, belonging to the field of intelligent waste material identification technology. The method involves acquiring visible light image data, depth structure data, and near-infrared spectral response data of the waste materials to be identified; obtaining candidate regions from the visible light image data and extracting a set of morphological features based on the depth structure data; extracting material reflection curves from the near-infrared spectral response data; extracting texture distribution features and color stability features from the visible light image data and generating a set of structural consistency features based on the morphological features; constructing a fusion feature vector by performing cross-modal association mapping on multiple features and inputting it into a waste material identification model to obtain candidate identification results; identifying key feature subsets through local sensitivity analysis and reconstructing the weights of the fusion features; and performing classification again to output the final identification result. This invention can effectively improve the accuracy and stability of waste material identification in complex recycling scenarios.

The application provides a kind of animation style virtual person animation real-time generation method and system, comprising: extracting multidimensional style features from animation reference map, wherein multidimensional style features include line drawing features, color features, texture features and character form features;From real-time captured portrait video, extract portrait key point sequence and interactive object sequence, and generate dynamic semantic identity anchor point stream according to interactive object sequence;With portrait key point sequence as structural constraint condition, with multidimensional style features as global guidance, dynamic semantic identity anchor point stream is embedded into latent space as hidden variable constraint, and controlled diffusion reasoning is carried out in latent space, to obtain animation latent feature frame sequence;Three-dimensional causal decoding and time domain smoothing filtering are carried out on animation feature frame sequence, to obtain smooth animation frame sequence;Real-time style feedback and style correction are carried out on smooth animation frame sequence, to obtain virtual person animation.

The invention discloses a hydraulic machinery stall vortex online identification method based on nonlinear dynamic characteristics, and relates to the field of hydraulic machinery operation monitoring and fault diagnose.The hydraulic machinery stall vortex online identification method can distinguish loading and unloading processes by introducing an empirical formula and a model of a guide vane change rate, accurately identifies a vortex strip state on a hysteresis loop, and improves the accuracy of stall vortex identification. The dynamic error is reduced by more than 80%; secondly, through an empirical formula, the input of the neural network has a clear hydraulics meaning (such as a velocity triangle relation), and the unreliability of a pure black box model is avoided; wherein the complex fluid physical law is concentrated in an empirical formula, the neural network only needs to process residual mapping, the operation speed is extremely high, and the method can be deployed in an in-situ control unit.

Provided are an analysis device, an analysis method, and a storage medium that can analyze the state of an electrode provided in a battery with high precision. The analysis device of the embodiment is an analysis device of an active material layer of an electrode of a secondary battery, which includes: an acquisition unit that acquires image data that represents an active material in the active material layer after ion milling processing and a gap or a filler between the active materials by using a difference in brightness; and a comparison unit that compares a pattern of the difference in brightness between the image data in at least two different states of the active material layer.

The invention relates to the technical field of geological disaster monitoring, in particular to a multi-source remote sensing and earth surface deformation fused frozen soil hot melt collapse identification system and method. The system comprises a multi-source data acquisition preprocessing module, a double-branch anti-interference fusion module, a WHF coupling evolution modeling module, a double-factor collaborative early warning module and a multi-dimensional visual output module. According to the method, specialized data extraction and calibration are realized by integrating multi-source remote sensing and ground data and constructing a double-branch architecture, the data purity is improved by adopting an anti-interference fusion algorithm, the limitation of single data monitoring can be broken through, meanwhile, a WHF coupling evolution model is constructed, a frozen soil phase change and evolution law is inverted, risk levels are scientifically divided in combination with a double-factor early warning mechanism, and the risk level of the frozen soil is improved. On the basis of a multi-dimensional visualization technology, monitoring results and risk distribution are visually presented, data export and report generation are supported, and efficient technical support is provided for disaster prevention and control in the permafrost region.

The application discloses a modified ResNet-50 ocean internal wave intelligent recognition method fusing multi-scale features, and belongs to the technical field of cross of marine satellite remote sensing and computer vision, and comprises the following steps: acquiring an original high-resolution remote sensing measured image, and obtaining an enhanced satellite remote sensing image through pretreatment; manually labeling and correcting the labeling discontinuity and adhesion problems; through morphological dilation, adaptive sliding window cropping, data enhancement and data set division, a balanced data set of positive and negative samples is constructed; a single-channel backbone network is reconstructed, an ASPP module is introduced to build a multi-scale feature network; an optimal detection model is obtained through iterative optimization of a coupled loss function; and a remote sensing image to be detected is input, multi-dimensional evaluation is carried out, and an internal wave pixel-level prediction mask is obtained. The application is suitable for single-polarization remote sensing image features, can suppress sea clutter interference, reduce internal wave false alarm and false alarm, and improve the internal wave recognition accuracy and topological restoration robustness under complex sea conditions.

The invention discloses a multi-mode medicine purchasing behavior data analysis system and method oriented to a smart pharmacy, and relates to the technical field of big data analys.The method comprises the steps that historical medicine purchasing records of a user are collected, complete medicine purchasing processes are constructed, the complete medicine purchasing processes are clustered, and a medicine purchasing reference database is constructed with the major function as a main index; after a user provides a prescription, generating a to-be-detected medicine purchasing process; retrieving a reference database according to the major function of the medicine, comparing the medicine positions of the to-be-detected process and the reference process for compliance verification, and delivering the medicine out of storage if the medicine is qualified; when the verification is not qualified, reversely comparing and screening the flow with the highest similarity, extracting the medicine in the corresponding order as a correction result, and manually checking the correction result; the system comprises a medicine purchasing record processing module, a reference database construction module, a to-be-detected process generation module, a medicine purchasing process verification module and a process correction checking module. The method can reduce drug identification errors, effectively intercept drug purchase abnormity, and improve drug purchase accuracy and efficiency.

This invention discloses an intelligent fault identification method and system for coal mines, relating to the field of coal mining. The intelligent fault identification method for coal mines includes the following steps: S1. Data loading; S2. Data preprocessing; S3. AI model construction and training; S4. Intelligent fault identification; S5. Optimization of identification results; S6. Result output and correlation. This intelligent fault identification method for coal mines establishes a multi-level cross-judgment system of primary core judgment factors + secondary auxiliary judgment factors, dynamically adjusting depth and angle thresholds to avoid misjudgment based on single data; through parameter correction, the deviation rate between the final fault parameters and the actual exposure is reduced.

The invention discloses an intelligent bag taking method and system based on target recognition analysis, and relates to the technical field of target recognition. The method comprises the following steps of bag taking interference analysis, bag taking identification analysis and bag taking decision feedback. According to the method, the bag taking interference analysis result is obtained based on the bag taking identification interference parameters so as to judge whether to carry out image dynamic acquisition adjustment or not, if yes, bag taking dynamic identification analysis is carried out after the image dynamic acquisition adjustment is carried out, and if not, bag taking dynamic identification analysis is directly carried out; recognition errors caused by ambient light changes or other factors are effectively avoided, recognition of the bag taking action is more accurate, whether a bag taking decision feedback mechanism is triggered or not is judged after bag taking dynamic recognition analysis is conducted, if yes, bag taking decision feedback is executed, and if not, image dynamic recognition optimization is executed, so that the bag taking action is more accurate. The reliability of the automatic bag taking machine in actual operation is effectively improved, and then recognition of the bag taking process of the automatic bag taking machine is more accurately carried out.