155results about How to "Low resolution accuracy" patented technology

The embodiment of the invention relates to a retrieval question and answer method and device based on a large model, a storage medium and equipment, and relates to the technical field of natural language process.The method comprises the steps that thinking chain splitting is conducted on a target query question, multiple sub-question splitting results are obtained, and model prompt information is determined according to the multiple sub-question splitting results; performing retrieval recall in a preset index knowledge base on the basis of the multiple sub-question splitting results to obtain multiple retrieval results of different dimensions; generating a context content generation text according to the retrieval results of the plurality of different dimensions and the model prompt information; and inputting the context content generation text into a preset content generation large model to obtain a preliminary retrieval result, and performing fact verification on the preliminary retrieval result to obtain a target retrieval result. The accuracy of the target retrieval result is improved.

The invention provides a multi-source data fusion high-altitude falling object detection system and a high-altitude falling object detection method. The system comprises an acceleration sensor, a sound sensor, a processor and an alarm module. The acceleration sensor monitors the acceleration change of the target falling object and preliminarily judges an object falling event; when the sound sensor preliminarily judges that an object falling event occurs, a sound signal generated by a target falling object is monitored, and sound source positioning is achieved; the processor receives data of the sensor and verifies an object falling event again through a data fusion algorithm, so that misjudgment is avoided. In addition, the alarm module further generates alarm information including the falling time, the object type and the falling position to give an alarm. The system solves the problem that in the prior art, the accuracy of object falling detection through a single-type sensor is low.

The invention discloses a carbon emission monitoring method and device, and relates to the technical field of environment monitoring. The carbon emission monitoring method comprises the following steps: S1, data integrity evaluation; s2, evaluating data accuracy; and S3, performing optimization evaluation. According to the invention, the data integrity evaluation is carried out through the obtained first carbon emission monitoring data to obtain the integrity evaluation value, whether data integrity optimization is carried out is judged based on the integrity evaluation value, and then the data accuracy evaluation is carried out according to the obtained second carbon emission monitoring data to obtain the accuracy evaluation value. Whether data accuracy optimization is carried out or not is judged based on the accuracy evaluation value, finally, evaluation is carried out according to the sensor precision, the integrity evaluation value obtained after optimization and the accuracy evaluation value obtained after optimization to obtain an optimization evaluation value, and whether feedback is carried out or not is judged based on the optimization evaluation value. The effect of improving the data accuracy of carbon emission monitoring is achieved, and the problem that in the prior art, the data accuracy of carbon emission monitoring is low is solved.

The invention relates to the field of geophysical exploration, and discloses a seismic data noise suppression method, which comprises the following steps: screening adjacent shot interference from an original single shot record by using a tau-p-q domain sparse inversion method to obtain a primary single shot record; linear interference is screened out from the primary single-shot record in a frequency domain transformation mode, and a secondary single-shot record is obtained; extracting an energy threshold value of the secondary single-shot record by utilizing a frequency division energy statistics mode, and performing primary noise suppression on the secondary single-shot record by utilizing the energy threshold value to obtain a de-noised single-shot record; and respectively extracting a primary effective signal and a surface wave signal from the de-noised single-shot record, extracting a standard single-shot record from the primary effective signal and the surface wave signal by using a frequency division noise automatic identification and attenuation method and an adaptive spectrum editing method, and ending noise suppression. The invention further provides a noise suppression device and equipment for seismic data, a medium and a program. According to the invention, the accuracy of seismic data imaging can be improved.

The invention relates to the technical field of vehicle carbon emission prediction, and particularly provides a digital twinning-assisted intelligent network-connected vehicle carbon emission prediction system and method, which combines digital twinning and deep learning, applies the combined digital twinning and deep learning to a carbon emission prediction link, and simulates various driving behaviors to obtain a carbon emission prediction result. And finally, giving an optimal carbon emission reduction mode based on the current driving state according to comparison, and feeding back the optimal carbon emission reduction mode to a user. Meanwhile, the whole prediction part can continuously compare a prediction result with an actual execution result to optimize the whole structure, so that the prediction accuracy and generalization are improved, and the user requirements are met.

The invention relates to the technical field of elevator control, and discloses an elevator task control method and system based on medical logistics robot scheduling. The method comprises the steps that according to a dispatching task of a medical logistics robot, the robot position and electric quantity information of an available medical logistics robot, a dispatching task to be executed, the floor where an elevator of an available elevator is located, the running direction, whether the elevator is fully loaded or not and the floor where the elevator needs to stop are recognized, and a distribution combination is constructed; an evaluation function is constructed according to the robot path length of each scheduling task in the distribution combination and the time of taking the corresponding available elevator, and a target medical logistics robot and a target available elevator are obtained through scheduling optimization based on the evaluation function; constructing a target control signal according to the target medical logistics robot and the target available elevator; and a scheduling path of the target medical logistics robot is generated, and elevator task control is conducted on the target available elevator according to the scheduling path and the target control signal. The accuracy of elevator task control can be improved.

The invention discloses an anomaly detection method and device for main network equipment and electronic equipment. The method comprises the following steps: acquiring historical operation data of main network equipment, and determining standard operation data of the main network equipment; inputting the historical operation data into the target model to obtain predicted operation data of the target date; calculating a historical deviation value based on the historical operation data and the standard operation data, calculating a predicted deviation value based on the predicted operation data and the standard operation data, and calculating a comprehensive deviation value through the historical deviation value, the predicted deviation value, the historical operation data, the standard operation data and the predicted operation data; and when the comprehensive deviation value is greater than or equal to a deviation value threshold value, determining the operation state of the main network equipment as an abnormal operation state, and when the comprehensive deviation value is smaller than the deviation value threshold value, determining the operation state of the main network equipment as a normal operation state. Through the method and the device, the problem of low accuracy of the anomaly detection result of the main network equipment in related technologies is solved.

The invention relates to a video abstract generation method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining a plurality of frames of video key images corresponding to key part features and key motion features in video data; respectively inputting the video key images into a plurality of pre-trained first network models to obtain corresponding prediction results; simultaneously extracting spatio-temporal information from multiple frames of video key images by using a first network model, and converting each frame of video key image into two-dimensional image features related to a time sequence; and combining the plurality of prediction results to obtain a video abstract of the video data. Through the stage recognition method and device, the problem of low accuracy of stage recognition in related technologies is solved, complementary information of key images of different frames of videos is provided by using two-dimensional image features and spatio-temporal information related to a time sequence, and then the accuracy of stage recognition is improved.

The invention discloses a detection method and device of an air insulation switch cabinet, a storage medium and electronic equipment, and relates to the technical field of air insulation switch cabinet state detection. The method comprises the following steps: detecting whether characteristic gas exists in the air-insulated switch cabinet according to type information of gas in the air-insulated switch cabinet; if the characteristic gas exists in the air-insulated switch cabinet, determining that a target defect exists in the air-insulated switch cabinet, and acquiring gas concentration information in the air-insulated switch cabinet; and inputting the gas concentration information in the air-insulated switch cabinet into the target evaluation model for processing to obtain the defect type and defect degree of the target defect in the air-insulated switch cabinet. According to the invention, the problem of low accuracy of detecting whether defects such as partial discharge and partial overheating exist in the air insulation switch cabinet in the prior art is solved.

The invention discloses an instrument panel data determination method and device, a storage medium and an electronic device.The method comprises the steps that an inspection image from a user interaction interface is received, and the inspection image is obtained after image collection is conducted on an inspection area; determining an identification result obtained by performing image identification on the inspection image by using a target identification model, and determining an instrument image of an instrument panel in the inspection area from the identification result; and determining a target instrument program corresponding to the instrument type of the instrument panel from preset instrument programs, and performing image processing on the instrument image according to an image recognition algorithm corresponding to the target instrument program to obtain instrument panel data of the instrument panel. By adopting the technical scheme, the problem that the accuracy of the instrument panel data is relatively low when the instrument panel data is automatically read is solved.

The invention discloses a multi-agent hyperspectral band selection method based on a fusion state and a superior experience pool, and belongs to the field of remote sensing image processing. After a hyperspectral image is obtained, a sample set of the hyperspectral image is generated, and the sample set is divided into a training sample set and a test sample set; a multi-agent waveband selection model is constructed according to the training sample set, the state representation method comprises meta description statistics, an auto-encoder and a graph convolutional network, and the reward allocation strategy comprises an adaptive allocation strategy based on waveband importance; inputting the training sample set into a multi-agent waveband selection model for model training to obtain a trained multi-agent feature selection model, and optimizing a memory sample of an experience pool for model training by adopting a specific sample oriented experience pool optimization method based on a Gaussian mixture model; and inputting the test sample set into the trained multi-agent feature selection model for band selection, outputting a hyperspectral image band set, and classifying the hyperspectral image band set to obtain a classification result.

The invention provides an Internet of Vehicles user voice emotion model optimization method and device and electronic equipment, and relates to the field of Internet of Vehicles user voice emotion recognition. Specifically, the method comprises the following steps: acquiring original voice text data for training a first voice emotion model, and first voice text data collected in an emotion prediction process by applying a second voice emotion model, the second voice emotion model is obtained by training the first voice emotion model through the original voice text data; screening out first target data from the first voice text data through the second voice emotion model; and optimizing the second speech emotion model by using the original speech text data and the first target data to obtain a target speech emotion model, the target speech emotion model being used for performing emotion prediction on the speech text data of the Internet of Vehicles user.

The invention provides a method and device for improving user intention recognition accuracy and electronic equipment. The method comprises the following steps: acquiring a current user question, extracting keywords of the current user question to obtain question keywords, and matching the question keywords with keywords in a knowledge base to obtain a keyword matching result; converting the question keyword into a question word vector, converting the keyword in the knowledge base into a knowledge base word vector, respectively determining the similarity and correlation between the question word vector and the knowledge base word vector in the knowledge base, and obtaining a similarity sequence and a correlation sequence of the question word vector; a standard question set in a knowledge base corresponding to the current user question is determined at least according to the similarity sequence, the correlation sequence and the keyword matching result, the user intention is determined according to the standard question set, and the standard question set comprises a plurality of standard questions. The problem that in the prior art, the accuracy rate of user intention recognition is low is solved.

The present invention relates to the field of intelligent interaction technology, and discloses a big data driven digital human intelligent interaction method and system, the method comprising: extracting interactive audio and video frame sets in user historical interactive videos, performing image smoothing and image equalization on each video frame in the video frame set to obtain enhanced video frames; identifying interactive texts corresponding to interactive audios; extracting feature time sequence information of enhanced video frames, interactive texts and interactive audios respectively, performing feature interaction on the feature time sequence information to obtain multimodal features; calculating target interactive texts and corresponding target interactive voices according to multimodal features, generating target interactive expressions according to target interactive texts and target interactive voices; constructing interactive videos of target digital humans according to target interactive texts, target interactive voices and target interactive expressions, and using interactive videos to interact with users. The present invention can improve the accuracy of intelligent interaction of digital humans.

The present invention relates to the field of digital information technology, and discloses a method and system for dispatching and allocating firefighting resources in a building based on BIM data, including: fusing BIM data and sensor data to obtain multi-dimensional fusion data; constructing a three-dimensional building model of a target building according to the multi-dimensional fusion data, simulating a fire on the three-dimensional building model, and obtaining a fire data image; extracting fire factor features of the fire data image, generating an evacuation path of the target building according to the fire factor features, allocating firefighting resources according to the evacuation path, and obtaining a firefighting resource allocation vector of the evacuation path; calculating a resource allocation score of the firefighting resource allocation vector; when the resource allocation score is greater than a preset resource allocation threshold, generating a firefighting resource dispatching strategy according to the firefighting resource dispatching vector, and allocating firefighting resources according to the firefighting resource dispatching strategy. The present invention can improve the accuracy of dispatching and allocating firefighting resources in a building.

The present application discloses a method and device for determining the implantation position of a steel plate based on the distal end of the femur. The method relates to the fields of data processing technology, medical device technology and other related technical fields. The method includes: performing three-dimensional reconstruction based on the image data of the distal end of the femur of the target object to obtain a three-dimensional model 1 corresponding to the distal end of the femur of the target object; determining a preset implantation area of ​​the target implanted steel plate based on the three-dimensional model 1; within the preset implantation area, implanting each of a plurality of steel plates to be implanted into the three-dimensional model 1 to obtain a plurality of three-dimensional models 2; determining the target implanted steel plate from a plurality of steel plates to be implanted based on the three-dimensional model 1 and the plurality of three-dimensional models 2, and determining the implantation position corresponding to the target implanted steel plate as the target implantation position. Through the present application, the problem that the implantation position of the medial steel plate at the distal femoral fracture is determined based on artificial experience in the related art, resulting in a relatively low accuracy in the planning of the steel plate implantation, is solved.

The application discloses a data processing method and device, a storage medium and an electronic device. It relates to the technical field of medical treatment. The method comprises the following steps: a tibia three-dimensional model is constructed, and a plurality of target steel nail three-dimensional models are obtained; a plurality of second medical images are obtained, and a plurality of target steel nail first two-dimensional pixel point sets and a tibia second two-dimensional pixel point set are obtained according to the plurality of second medical images; target space conversion parameters of the plurality of target steel nails in the tibia coordinates are obtained by calculation according to the target steel nail first two-dimensional pixel point sets, the tibia second two-dimensional pixel point set, the tibia three-dimensional model and the steel nail three-dimensional models; a guide plate model is obtained according to the target space conversion parameters, and a target space posture of a tibia osteotomy plane is obtained according to the guide plate model. Through the application, the technical problem that the space posture of the tibia osteotomy plane is relatively low in accuracy because the space posture of the tibia osteotomy plane is determined by relying on artificial experience is solved.

The present invention discloses a text processing method and device. The method comprises: extracting erroneous words from text information, wherein the erroneous words include at least one of the following types: grammatical errors, vocabulary errors, and spelling errors; obtaining correction words corresponding to the erroneous words based on word collocation information of the field to which the text information belongs, wherein the word collocation information is obtained based on the corpus of the field to which the text information belongs; and correcting the text information based on the correction words to obtain the corrected text information. The present invention solves the technical problem in the prior art that the accuracy of the trial record obtained by voice conversion of the recording during the trial is low.

The invention relates to a user data processing method and system, a storage medium and electronic equipment, and relates to the technical field of big data process.The method comprises the steps that data cleaning is conducted on original user attribute data of to-be-classified users to obtain target user attribute data, the to-be-classified users are subjected to bucket dividing according to the target user attribute data, and the buckets are obtained; obtaining a user bucket dividing result; determining an attribute association relationship between user entities corresponding to the to-be-classified users included in the same user bucket dividing result, and constructing an entity triple according to the attribute association relationship and the user entities; determining collaborative association edges among the user entities according to the entity triples, and constructing a user association feature map based on the collaborative association edges, the user entities and the target user attribute data; and inputting the user association feature map into a preset user classification model to obtain a user classification result, and determining the user category of the to-be-classified user according to the user classification result. The accuracy of the user classification result is improved.

The present invention relates to the technical field of electronic non-stop toll collection, and in particular to an electronic non-stop toll collection system based on high-precision identification, including a data acquisition module, an information processing module, and an identification and deduction module. The present invention greatly improves the accuracy of license plate recognition and the reliability of the system through advanced data acquisition and processing technology. The system collects license plate images through image sensors in the environmental monitoring area, and uses character recognition and judgment units for high-precision recognition to ensure that a high recognition rate can be maintained even under adverse weather conditions. By real-time monitoring of vehicle speed and image clarity, the system can dynamically adjust the heating temperature to prevent image blur and equipment failure. The system automatically adjusts system parameters for interfering weather such as rain and snow, thereby improving the stability of license plate recognition under various climatic conditions. It effectively solves the problem of low accuracy of license plate information recognition under extreme weather conditions due to reliance on light sources and light sensors.

The invention relates to a case test result processing method and device, a storage medium and electronic equipment, and relates to the technical field of software test.The method comprises the steps that a case execution result obtained by executing a test case is obtained, and a case name and a case error reporting state of the test case are extracted from the case execution result; classifying the use case execution results based on the use case names and the use case error reporting states to obtain use case classification results; extracting the use case attribute information of each test case included in the same use case classification result, and comparing the use case attribute information of each test case in the same use case classification result to obtain an attribute comparison result; and combining the test cases included in the same case classification result according to the attribute comparison result to obtain a case combination result. The processing efficiency of the use case test result is improved.

The invention relates to the technical field of computer vision, and provides a pedestrian re-recognition model training method and device, electronic equipment and a storage medium. The method comprises the following steps: inputting a training image into a student model, performing feature extraction on the training image, and performing random intensity shielding on a local feature vector obtained by feature extraction to obtain an enhanced feature vector; inputting the enhanced feature vector and the first initial category feature vector into a feature prediction network of a student model to obtain a first feature vector and predicted shielding intensity of the sub-image; inputting the training image into a teacher model to obtain a second feature vector; and monitoring the first feature vector by using the second feature vector, predicting the shielding intensity by using the random shielding intensity monitoring sub-image, updating the parameter of the student model and the parameter of the teacher model, and determining the trained teacher model as the pedestrian re-identification model, thereby improving the performance of the pedestrian re-identification model under the shielding condition. And the robustness and generalization ability of the pedestrian re-identification model are enhanced.

The present application discloses a method, device, storage medium and processor for executing a work order task. The method comprises: receiving target information in a task message of a work order in a manufacturing production management process, wherein the target information comprises at least one of the following: primary key identification information of the task message, message queue information of the task message; parsing the target information to obtain the target primary key identification information; determining the target message queue information corresponding to the target primary key identification information, wherein the target message queue information comprises a plurality of ordered task messages; and executing the target execution strategy of the task message according to the target message queue information. Through the present application, the problem of low accuracy in the execution of work order tasks in the related art is solved.

The invention discloses a rehabilitation device for preventing foot drop of a hemiplegic patient, the rehabilitation device comprises a sole supporting plate and a shank supporting plate, the heel part of the sole supporting plate is movably connected with the bottom of the shank supporting plate through a first spherical shaft device, and one side of the sole supporting plate is provided with a second spherical shaft connecting device; a third spherical shaft connecting device is arranged on the other side of the second spherical shaft; by combining a plurality of structures such as the sole supporting plate, the shank supporting plate, the first spherical shaft device, the second spherical shaft connecting device and the third spherical shaft connecting device, the angle between the sole supporting plate and the shank supporting plate is accurately adjusted through a computer program stored on a computer readable storage medium; according to the technical scheme, ROM training of double-lower-limb massage training can be carried out on a paralyzed patient, and according to the training basis, the product is used for assisting the patient in carrying out joint flexion and joint stretching activities of feet, so that joint capsules and tendons around joints are fully stretched.

The present disclosure provides a text detection method, electronic device, medium and computer program product, which relate to the field of data processing technology. The method includes: obtaining image features of a target object; inputting the image features into a first decoder of a text detection model to obtain a first feature, and obtaining a first text detection result based on the first feature; inputting the first feature into a second decoder of a text detection model to obtain a second feature, and obtaining a second text detection result based on the second feature; wherein the first text detection result and the second text detection result are text detection results of different levels, and the levels of the text detection results include: word level, line level, or paragraph level. The present invention can obtain multiple levels of text detection results at the same time, and does not need to perform text detection on another level based on the text detection result of one level, but obtains the text detection results of each level based on the first feature, and the text detection results are highly accurate.

This invention provides a method, apparatus, storage medium, and electronic device for adjusting weights. The method includes: scoring a target image from multiple preset dimensions, determining a dimension score for each preset dimension based on the scoring results and the weight of each preset dimension, and determining a deviation score between the target image and a reference image; determining a target comprehensive score for the target image based on the dimension scores and the deviation scores; repeating the following operations if the target comprehensive score does not meet a first condition until the re-determined target comprehensive score meets the first condition: adjusting the dimension weights of the target dimensions included in the multiple preset dimensions, re-determining the dimension scores based on the scoring results and the adjusted dimension weights of the target dimensions, and re-determining the target comprehensive score based on the re-determined dimension scores and the deviation scores.

The invention discloses a data processing method and device, a storage medium and electronic equipment. The method comprises the following steps: constructing a tibia three-dimensional model, and obtaining a steel nail three-dimensional model of a plurality of target steel nails; obtaining a plurality of orthogonal second medical images, and obtaining a first two-dimensional pixel point set of the plurality of target steel nails and a second two-dimensional pixel point set of the tibia according to the plurality of second medical images; calculating according to the first two-dimensional pixel point set, the second two-dimensional pixel point set, the tibia three-dimensional model and the steel nail three-dimensional model to obtain target space conversion parameters of the multiple target steel nails under the tibia coordinates; and obtaining a guide plate model according to the target space conversion parameters, and obtaining a target space attitude of the tibia osteotomy plane according to the guide plate model. By means of the method and device, the technical problem that the accuracy of the space posture of the tibia osteotomy plane is low due to the fact that the space posture of the tibia osteotomy plane is determined through related technologies depending on artificial experience is solved.

The invention provides a crop growth vigor identification method and device. The method comprises the following steps: performing block processing on a crop height matrix and a land height matrix to obtain a plurality of small crop matrixes and a plurality of small land matrixes; calculating longitudes and latitudes corresponding to the small crop matrixes and the small land matrixes; determining a small crop matrix and a small land matrix corresponding to the same land grid in the target land according to the longitude and latitude corresponding to each small crop matrix and each small land matrix; calculating a difference value between the small crop matrix corresponding to each land grid and the small land matrix to obtain a crop net height matrix corresponding to the land grid; and based on the crop net height matrix corresponding to each land grid, identifying an area with non-uniform growth vigor of crops in the target land. By adopting the technical means, the problem that in the prior art, the accuracy of identifying whether the growth vigor of the crops in the complex terrain is uniform or not is low is solved.

The invention provides an image voiceprint multi-source perception data cross mutual verification method and device and electronic equipment. The method comprises the steps that a data feature set of each data channel in a current observation time window is acquired, fault identification is carried out on the electrical equipment according to the data feature set, a fault identification result is generated, and the fault identification result comprises a fault data sequence of each data channel; obtaining a target fault data sequence of the target data channel from the fault recognition result, and determining other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the starting time of the target fault data sequence; and determining a cross mutual verification result of the target data channel at least according to the target fault data sequence and other fault data sequences, and determining an equipment fault identification result of the target data channel according to the cross mutual verification result. The problem that a traditional power distribution network equipment fault identification cross-checking method is low in fault identification accuracy is solved.

The invention provides a noise acquisition device, which comprises an audio acquisition device used for acquiring audio data of a to-be-tested motor in a sound insulation environment; the posture adjusting mechanism is used for bearing and adjusting the posture of the to-be-tested motor and changing the gravity center position of the to-be-tested motor in the testing process so as to detect noise changes of the motor at different angles, and therefore problems possibly existing in the motor can be recognized; and the excitation module is used for applying external excitation to the to-be-tested motor in the testing process to promote impurities or defects in the motor to generate sound, so that the quality problem is exposed. According to the device, through a comprehensive test method combining sound collection, attitude adjustment and internal excitation, accurate collection and analysis of motor noise can be realized, manpower is effectively replaced, the detection efficiency and accuracy are improved, and the device is suitable for noise quality detection of the motor under different attitudes.