36 results about "Subjective perception" patented technology

The invention discloses a college dormitory noise multi-dimensional evaluation method based on psychoacoustic and sound field simulation, and belongs to the field of building environment and noise control. The objective of the invention is to solve the problem that an existing dormitory noise evaluation method is single in index and neglects subjective perception and spatial distribution heterogeneity. The method comprises the following steps: acquiring physical data and original audio of dormitory noise through multi-point synchronous acquisition; calculating loudness, sharpness, roughness, fluctuation intensity and comprehensive psychoacoustic annoyance (PA) by using a psychoacoustic model, and scientifically quantifying subjective interference of noise; meanwhile, sound field simulation is carried out based on the accurately established three-dimensional dormitory model, a sound pressure level distribution cloud picture is generated, and a standing wave phenomenon and a sound energy accumulation area are visually revealed; and finally, generating a targeted noise reduction optimization strategy by integrating subjective and objective and spatial analysis results. According to the method, full-chain diagnosis of physical characteristics-subjective perception-spatial distribution of dormitory noise is realized, and a scientific basis is provided for accurate and efficient noise treatment.

Aircraft high-load landing event detection, classification and maintenance guidance methods relate to the technical field of aircraft high-load landing event detection. The aircraft high-load landing event detection method collects complete and accurate landing data online during an aircraft landing process and performs an analysis and calculation to determine whether a high-load landing event occurred and determines the type of the high-load landing event. The aircraft high-load landing event detection method removes the high-load landing event determination from the pilot's subjective perception and does not need to wait for decoding of quick access recorder (QAR) data, greatly improving the accuracy and timeliness of high-load landing event detection. In addition, by analyzing and estimating structural loads of the aircraft, the aircraft high-load landing event detection method provides targeted maintenance recommendations for possible aircraft structural damage caused by the high-load landing event, and forms a corresponding report.

The invention discloses an urban park perception value evaluation method based on multi-modal data fusion, and the method comprises the steps: extracting a multi-dimensional perception tag and an emotion intensity score from park social media text comment data through a pre-trained language model, and carrying out the aggregation of the multi-dimensional perception tag and the emotion intensity score, and generating a subjective perception result; performing element identification on the image comment data of the park social media by using the three models to obtain different scene element coverage rates, character / facility counts and language representations of visual elements; constructing a plurality of objective environment indexes to describe different environment states, extracting variable information from three model output results, POI data, remote sensing land class proportion and road network information, and constructing mapping with the corresponding objective environment indexes to quantify each objective environment index; and aligning and outputting the subjective perception result and the objective environment index in the park and period dimensions. According to the method, automatic quantitative evaluation of subjective and objective characteristics of urban parks can be realized on the national scale, and standardized technical support is provided for urban renewal and ecological management.

The invention discloses an earphone defect detection method and device based on voiceprint feature analysis and a storage medium, and relates to the technical field of electroacoustic quality control. According to the method, human ear hearing characteristics are simulated, Mel-cepstrum voiceprint characteristics are extracted from sound response signals, an adaptive weight mechanism based on an equal-loudness curve is introduced, and a voiceprint abnormal deviation index model conforming to human ear subjective perception is constructed; in order to solve the problem that tiny abnormal noise is difficult to recognize, high-order statistics such as kurtosis and skewness are adopted for feature fusion enhancement, and a support vector machine is combined to achieve accurate capture of transient impulse noise. According to the invention, various defect types such as voice coil friction, imbalance of balance degree, diaphragm damage and the like can be fully automatically distinguished, the problems that a traditional test means is disjointed with subjective hearing sense and the omission ratio of hidden defects is high are effectively solved, and the quality inspection efficiency and accuracy of earphone production lines are remarkably improved.

This invention relates to the field of big data analytics, specifically to a method and system for intelligent analysis and assessment of risk perception bias among residents in post-disaster reconstruction areas. The method involves fitting risk indicators using multi-source data and a collaborative model, and calculating risk assessment values ​​based on the fitting results. It obtains basic information data, subjective perception data, and resident behavior data through structured questionnaires. Initial correction coefficients are matched to the subjective perception data based on the basic information data, and error analysis is performed using resident behavior data to determine whether dynamic compensation of the initial correction coefficients is necessary. This significantly improves the accuracy of the final obtained true subjective perception data. By assessing residents' risk perception bias from two perspectives—true subjective perception data and risk assessment values—the method ensures the accuracy and comprehensiveness of the final results, achieving intelligent analysis and accurate assessment of risk perception bias among residents in post-disaster reconstruction areas.

The invention relates to the technical field of physiological index data processing, in particular to a user feedback driven physiological index adaptive health assessment method and device. The method comprises the following steps: acquiring a BCG signal of a user based on a non-contact sensor, and acquiring index monitoring data of a plurality of indexes based on the BCG signal; the user inputs a user feedback text representing the abnormal state of the body, matching is conducted on the basis of the user feedback text and the disease words in the lexicon, and if the disease words are matched, indexes corresponding to the disease words serve as final monitoring indexes; if the disease word cannot be matched, screening out an initial monitoring index based on the user feedback text, and determining a final monitoring index based on the initial monitoring index; and obtaining daily index monitoring data based on the final monitoring index. In this way, mapping from coarse-grained subjective perception to refined objective indexes can be realized, and monitoring blind areas and misjudgment risks are remarkably reduced while the early-stage anomaly detection rate is increased.

The application discloses a kind of urban park perception value evaluation methods based on multi-modal data fusion, utilize pre-trained language model to park social media text comment data extraction multidimensional perception label and sentiment intensity score, and subjective perception result is aggregated and generated;Park social media image comment data is identified using three kinds of models, obtain different scene element coverage, figure / facility count, and language representation of visual elements;A plurality of objective environment indexes are constructed to describe different environmental conditions, variable information is extracted from the output results of the three models, POI data, remote sensing land class proportion and road network information, and a corresponding objective environment index is constructed to map to quantify each objective environment index;Subjective perception result and objective environment index are aligned in park and period dimension and output.The present application can realize the automatic quantification evaluation of subjective and objective characteristics of urban park at national scale, and provide standardized technical support for urban renewal and ecological management.

The invention relates to a video restoration method and device based on bullet screen feedback, equipment and a medium, and the method comprises the steps: collecting bullet screen data in a video playing process, and recognizing feedback data used for feeding back a video quality problem from the bullet screen data, the feedback data comprising a bullet screen text and bullet screen time; determining a problem type and a problem severity degree of a video quality problem according to the bullet screen text, and determining a problem region of the video quality problem according to the bullet screen text and the bullet screen time; and determining a repair strategy according to the question type and the question severity, and performing video repair on the question region according to the repair strategy. Through a bullet screen semantic recognition mechanism, subjective feedback information of a user on image quality in a video watching process is converted into quantifiable quality problem description, information mapping from subjective perception to objective restoration is realized, and the problem that faults exist between subjective quality feedback of the user and an objective video restoration technology during video restoration is solved.

The invention relates to an image compression method and system based on large and small model collaborative driving, and the method comprises the steps: carrying out the preprocessing of an input image, extracting the image features of each image block of the input image, and carrying out the quantification; side information of the image features is extracted, quantized and compressed into a side information code stream; decoding the side information code stream, and performing feature transformation on the quantized side information to obtain super-prior information; inputting the quantized image features and the hyper-prior information into an entropy model to obtain probability distribution parameters of the quantized image features; compressing the quantized image features into a main code stream, and predicting according to probability distribution parameters to obtain the size of the main code stream; and decoding the main code stream according to the size of the main code stream to obtain reconstructed quantized image features, and realizing multi-target output through a main decoder. Compared with the prior art, the objective signal fidelity of the decoded image in a high code rate interval, the subjective perceptibility of the decoded image in a low code rate interval and the analysis performance of the compressed image in a machine vision scene are remarkably improved.

The invention discloses a Mini-LED backlight display halo perception boundary determination method and system, and aims to solve the problem that a halo boundary in a high dynamic contrast display scene is difficult to objectively define. The method comprises the following steps: acquiring brightness distribution of a display area and a surrounding dark area in a dark environment; guiding an observer to determine a subjective halo boundary position under a standard observation condition; mapping the position to objective brightness data to extract a boundary brightness threshold value; and establishing a sensing boundary model based on multiple groups of peak brightness and boundary thresholds, wherein the sensing boundary model is used for stably determining a halation region which can be sensed by human eyes under different brightness, partition and dimming strategies. According to the method, the perception boundary is used for replacing a fixed absolute threshold value, evaluation consistency is remarkably improved, closed-loop association of subjective perception and objective data is achieved, and reproducible and computable standardized criteria are provided for halo evaluation, algorithm optimization and production line quality control.

The application discloses a kind of based on modular design's text video quality evaluation method and system, belong to computer vision technical field.The method includes: extracting time sequence, space and the multi-dimensional feature of graphic-text matching degree from original text video;Initial quality score of video is generated using multi-modal base model;Multi-dimensional feature and initial score are fused to generate prediction label;Through feature mapping block generation weight and bias coefficient, initial score is adaptively corrected to obtain final quality score;Using minimum batch learning strategy, combine prediction label dataset with real label dataset, to train quality evaluation model in a supervised learning manner.The application realizes the effective fusion of multi-dimensional quality characteristics through modular design, significantly reduces the dependence of the model on artificial annotation data, improves the consistency of evaluation results and human subjective perception, and provides an efficient and reliable solution for text video quality evaluation.

The invention provides an image quality evaluation method and device and a medium, and belongs to the technical field of video image processing, and the method comprises the steps: obtaining a to-be-evaluated image; dividing the to-be-evaluated image into a plurality of areas; determining each evaluation index of the plurality of areas according to a preset weighting rule to obtain a plurality of evaluation information; and fusing the evaluation information to obtain a quality evaluation result of the to-be-evaluated image. According to the method, the image is subjected to region division, and multi-level and multi-dimensional analysis and quantification are carried out, so that the final quality evaluation result can more accurately reflect the perception quality of human eyes on the image, and the problem of disjunction with subjective perception caused by global statistics or average calculation in a traditional quality evaluation method is solved; and the subjective consistency and the evaluation efficiency of the image quality evaluation result are improved.

The invention provides an image quality evaluation method and device, and relates to the technical field of image processing. The method comprises the steps of obtaining a to-be-evaluated image; performing semantic feature extraction on the to-be-evaluated image by using a Mamba model to obtain multi-scale content features, the multi-scale content features including local distortion detail information and global content information; generating a weight parameter and an offset parameter of a target network by adopting a super network based on the multi-scale content features; and based on the multi-scale content features and the weight parameter and the bias parameter of the target network, adopting a preset image quality prediction target network to evaluate and obtain an image quality result. End-to-end blind image quality evaluation of global dependence modeling, local distortion capture and adaptive quality prediction can be realized, the accuracy of image quality evaluation is improved, and the deviation between an evaluation result and human subjective perception is reduced.

The invention discloses an image color aesthetics evaluation model and an image color aesthetics evaluation method based on a Kolmogorov-Arnold network. According to the method, firstly, multi-level color features of an image are extracted through a Kannform module integrated with a KAN, and modeling of a color relation is enhanced by using the powerful nonlinear capability of the multi-level color features; then, a multi-level feature aggregation module based on KAN space-time attention is adopted, color features of different levels are fused in a self-adaptive mode, and key color areas are highlighted; and finally, mapping the fused features into aesthetic scores by using a KAN feature regression device. The system comprises corresponding function modules. The problems that an existing method is insufficient in color space fine grit perception and prone to being interfered by non-color information, and the nonlinear mapping capacity of a traditional regression device is limited are solved, and the consistency, accuracy and stability of an evaluation result and subjective perception of human eyes are remarkably improved; the method can be widely applied to the fields of smart phone imaging evaluation, social media image screening and the like.

The invention discloses an abdominal tumor angiography magnetic resonance image generation method based on cross semantic perception, and the method comprises the steps: carrying out the refining of a feature graph through introducing a generator based on feature enhancement and multi-scale attention, and enhancing the expression capability of texture details and edge information; in order to enable a discriminator to pay more attention to the spatial and semantic features of the image, the discriminator based on cross semantic perception is designed, and the semantic features of the real angiography image are introduced into the discriminator to serve as additional guide information. Furthermore, the reversibility of conversion is ensured by utilizing cyclic consistency loss, the semantic difference of the image is learned through perception loss, more attention is paid to the visual quality of the image, and the image is closer to the subjective perception of human eyes on the image quality, so that the problem that pixel points between the generated image and the real image are not aligned is solved.

The invention discloses a video quality evaluation method and device, equipment and a storage medium, and belongs to the technical field of video processing. The method comprises the following steps: acquiring to-be-evaluated video data; extracting saliency semantic information corresponding to the visual attention area in the to-be-evaluated video data, and performing space-time modeling and context analysis on the saliency semantic information to obtain saliency features of the to-be-evaluated video data; extracting multi-scale image features of the to-be-evaluated video data; extracting coding domain features of the to-be-evaluated video data; fusing the saliency features, the multi-scale image features and the coding domain features to obtain fused feature representation; and predicting the fusion feature representation by using a pre-trained video quality evaluation model to obtain a quality evaluation result of the to-be-evaluated video data. According to the method, the video quality can be evaluated by integrating the multi-dimensional features, the problems that various distortion types are difficult to consider and subjective perception of human eyes is difficult to reflect in the prior art are effectively solved, and the video quality evaluation accuracy is improved.

The invention discloses a speech emotion recognition method based on subjective perception deviation knowledge distillation, and the method comprises the following steps: extracting acoustic features of a speech sample to be recognized and subjective perception deviation features corresponding to the speech sample, the acoustic features including spectrum features and self-supervised learning speech representation; inputting the acoustic features and the subjective perception deviation features into a teacher model for fusion to obtain fusion features, and generating a soft label based on the fusion features; inputting the acoustic features into a student model, processing the acoustic features through the feature fusion attention module to obtain student features, and generating a prediction label based on the student features; performing knowledge distillation on the student model by adopting the soft label, and obtaining a trained student model by minimizing the distribution difference between the prediction label and the soft label; and only inputting the acoustic features into the trained student model, and reasoning and outputting an emotion recognition result.

The invention relates to a multi-dimensional evaluation method for optimizing a content distribution strategy, and the method comprises the steps: constructing a comprehensive accelerated distribution effect comprehensive evaluation index system, and analyzing the performance of a whole network within a certain period of time after the distribution strategy takes effect. Quantitative evaluation is carried out from multiple aspects of request hit rate, user time delay, bandwidth consumption, storage occupation, cache updating and the like. The system not only pays attention to the subjective perception of the user in the resource access process, but also quantifies the actual conditions of user experience and system resource utilization through specific performance indexes. Through the multi-dimensional quantitative analysis, the subjective feeling of the user on the service experience can be converted into specific objective data, and a reliable basis is provided for scientific evaluation of the distribution effect.

The invention relates to a road real-time collision accident risk identification method and system driven by image data, and belongs to the technical field of road traffic safety. According to the method, subjective perception characteristics of a driver and objective traffic environment complexity are comprehensively considered, a deep learning model fusing the two interactive factors is provided, an SSM model is selected to describe attention distribution of the driver to an image, a DeepLabv3 model is adopted to mark semantic characteristics of the image, subjective and objective attributes of a feature vector fusion image are constructed, and the driver's attention distribution is obtained. And a gradient boosting decision tree algorithm is used to analyze the relationship between the various types of accident data and the feature vectors, a potential high-risk area is identified, and the accuracy of traffic accident risk identification is improved. The method has the technical characteristics of image driving, subjective and objective combination, high operability, low cost, real-time performance and the like, and is easy to popularize and apply.

The application provides a kind of urban space dynamic sound scene map construction method, system and medium, belong to the field of dynamic prediction, data integration, space mapping and visual display of urban sound scene, the method comprises: first, by fusing subjective perception score of residents and acoustic, ecological and visual objective environmental data, construct comprehensive sound scene perception index;Further, combined with acoustic simulation and measured correction technology, low-cost generation covers the high-density sound scene training sample set of global;Subsequently, the complex relationship between static space features and dynamic environmental features and sound scene perception index is learned using a machine learning model to establish a dynamic prediction model;Finally, the sound scene perception index prediction value output by the model is rendered through a GIS platform to generate a dynamic sound scene map that reflects the rhythm of urban operation. The present application overcomes the defects of existing sound scene technology such as static, single-dimensional and low resolution, and realizes dynamic, fine and interpretable quantification and visualization of human sound scene experience.

The invention provides an audio-visual quality evaluation method and system for audios and videos, electronic equipment and a storage medium, and relates to the technical field of audio-video evaluation, and the method comprises the steps: collecting the biometric data of a user in real time in the playing process of a target audio and video, and the environment data of a playing environment; performing feature extraction according to the biometric data to obtain biometric features, and converting the environmental data into environmental features; performing feature extraction on the target audio and video to obtain audio-visual signal features; performing multi-modal feature fusion on the biometric features, the environment features and the audio-visual signal features to obtain a fusion feature vector; and inputting the fusion feature vector into a pre-trained audio-visual quality evaluation model to obtain an audio-visual quality evaluation result of the target audio and video. According to the method, subjective perception and accurate prediction of the audio-visual quality are realized, the limitation that only signal indexes are relied on traditionally is broken through, and the accuracy and individuation level of audio-visual quality evaluation are remarkably improved.

The invention relates to the technical field of medical health monitoring, in particular to a liver, gall and pancreatic tumor patient symptom evaluation system based on multi-dimensional data. The data acquisition module is used for synchronously acquiring clinical diagnosis and treatment data of a patient, physiological index data, behavior activity data and medical image data of the wearable equipment and subjective perception data of the patient; the time sequence feature extraction module is used for performing feature extraction on the time sequence data by adopting an improved attention mechanism LSTM network to obtain a dynamic feature vector; multi-dimensional fusion innovation is achieved, a traditional single data evaluation framework is broken through, clinical, physiological, behavior, image and subjective perception data are fused for the first time, evaluation comprehensiveness is improved by 60% or above, an improved attention LSTM network is adopted to capture time sequence change characteristics of symptoms, meta-learning dynamic weight distribution is combined, the symptom evaluation accuracy rate reaches 89.3%, and the evaluation accuracy rate reaches 89.3%. Compared with a traditional method, the efficiency is improved by 25%.

The application discloses a kind of based on cross semantic perception's abdominal tumor contrast magnetic resonance image generation method, the generator based on feature enhancement and multiscale attention is introduced in the present application, and the expression ability of texture details and edge information is enhanced by refining feature map;In order to let discriminator can pay more attention to the spatial and semantic features of image, the discriminator based on cross semantic perception is designed, and the semantic features of real contrast image are introduced in the discriminator as additional guide information.Further, the reversibility of conversion is ensured by using cycle consistency loss, the semantic difference of image is learned by using perception loss, more attention is paid to the visual quality of image, and it is closer to the subjective perception of human eye to image quality, so as to solve the problem of pixel point misalignment between generated image and real image.

The invention discloses a text video quality evaluation method and system based on modular design, and belongs to the technical field of computer vision. The method comprises the following steps: extracting time sequence, space and image-text matching degree multi-dimensional features from an original text video; generating an initial quality score of the video by using the multi-modal basic model; the multi-dimensional features and the initial scores are fused to generate prediction labels; generating a weight and a bias coefficient through feature mapping partitioning, and performing self-adaptive correction on the initial score to obtain a final mass score; and training a quality evaluation model in a supervised learning mode by adopting a minimum batch learning strategy and combining a prediction label data set and a real label data set. According to the method, effective fusion of multi-dimensional quality features is realized through modular design, the dependence of the model on manual annotation data is remarkably reduced, the consistency of an evaluation result and human subjective perception is improved, and an efficient and reliable solution is provided for text video quality evaluation.

The invention discloses a non-reference image quality evaluation method based on mask reconstruction, and belongs to the technical field of image processing and computer vision. According to the method, a mixed synthesis image containing a distorted area and an undistorted area is constructed, and local controllable distortion is introduced under the condition that global semantics are kept consistent; on the basis, a semantic perception encoder based on mask reconstruction is used for extracting global semantic features insensitive to distortion, and a distortion perception encoder is used for extracting distortion features reflecting the image degradation degree under the constraint of the semantic features; meanwhile, a local degradation perception encoder is introduced to obtain fine-grained local degradation features; and finally, dynamically fusing the multiple types of features through an adaptive information fusion module, and outputting an image quality evaluation result. According to the method, image quality evaluation can be realized without a reference image, the consistency between a prediction result and human subjective perception can be effectively improved, and the method has good robustness and generalization ability.

This invention provides a method for assessing rural livability based on the national landscape engineering indicator system, relating to the field of assessment systems. The method includes: establishing a five-level rural livability indicator system and dividing the indicators into objective data anchors and subjective perception anchors; establishing objective data lakes and subjective data lakes to store corresponding data; receiving assessment focus trigger instructions and determining core indicators; recruiting relevant anchors in the indicator association network; capturing objective data sequences based on time ranges and retrieving or actively collecting subjective perception data to form data context packages; constructing a graph structure to assign data to indicator nodes and generate influence edges to instantiate an interaction graph; performing adversarial simulation on the graph through optimized agents and perturbation agents to generate a dynamic diagnostic report; and finally extracting correlation trends through pattern analysis and updating anchor collection priorities and protocols to achieve closed-loop optimization and self-evolution of the assessment process.

The invention discloses a video quality evaluation method and device, a storage medium and electronic equipment. The method comprises the following steps: inputting to-be-evaluated video data into a pre-trained video quality evaluation model; extracting a frame-level spatial feature corresponding to each video frame of the video data through a convolutional neural network in the video quality evaluation model; processing the frame-level spatial features through a time sequence model in a video quality evaluation model, and capturing a time dependency relationship between the video frames to determine spatial-temporal features corresponding to the video data; extracting semantic features corresponding to the video data through a visual language model in the video quality evaluation model; fusing the spatial-temporal features and the semantic features into video comprehensive features by using a fusion algorithm; and mapping the video comprehensive feature into a quality score corresponding to the video data through a regression device in the video quality evaluation model. According to the invention, comprehensive fusion from pixel-level features to semantic understanding is realized, and the accuracy of video quality evaluation in a complex scene and the consistency with human subjective perception are improved.

The application discloses a kind of based on VR interaction's immersion index calculation method, device, system and equipment, it includes obtaining the brain wave EEG data of multiple users, and subjective perception quality evaluation index IPQ, calculate the power spectral density of brain wave EEG data in preset frequency band, the IPQ obtained is combined with immersion dimension weighted conversion function, corresponding IPQ initial value is obtained, and the mean of the IPQ initial value corresponding to multiple field angles is calculated, to obtain the output value IPQ corresponding to dynamic virtual scene norm , build the calculation model of immersion index, obtain the first weight coefficient and the second weight coefficient, and determine the loss value of the calculation model using numerical difference method, obtain the first weight coefficient and the second weight coefficient when the loss value is minimum, and optimize the calculation model based on the first weight coefficient and the second weight coefficient when the loss value is minimum, by the above setting, improve the accuracy, comprehensiveness and objectivity of calculating the immersion index.

The application discloses a kind of based on children allergic rhinitis desensitization treatment tracking system, it is related to children allergic rhinitis desensitization treatment technical field, including: data acquisition module, for synchronously obtaining physiological signal data and discrete symptom evaluation data of target child in initial period of desensitization treatment;Signal processing module is used for the frequency domain analysis to physiological signal data, and the day rate heart rate variability index reflecting sympathetic and parasympathetic nervous balance degree of autonomous nervous system is calculated.This application establishes the dynamic calibration mechanism of objective physiological index to subjective symptom data, effectively eliminates the subjective perception deviation when physiological state is abnormal, so that symptom score can accurately reflect the real rhinitis inflammation load of child, effectively solve the traditional subjective and objective data static split, subjective data low reliability Problem, corrects data distortion problem, can reliably reflect real pathophysiological change, improve the accuracy of subsequent judgment result.

An exemplary system and method that can predict or estimate using machine learning based classification the onset of hot flash events or episodes that can be used to trigger the preemptive cooling intervention to prevent the experience of hot flash or to reduce the severity of the symptoms. The machine learning based classification can employ data from electrodermal activity (EDA) sensor measurements to generate a trigger for administration of active cooling via an active wearable device or garment prior to the subjective perception of hot flashes.