60 results about "Noise enhancement" patented technology

A method for diagnosing rotor rub-impact in a rotating machinery based on vibration signal deconstruction and frequency modulation characteristic anti-noise enhancement, including: converting, based on two-stage integration transformation and high-pass filtering, vibratory acceleration signals from a rotating machinery device into vibratory displacement signals; performing targeted extraction on a rotating frequency component therein based on an improved variational mode decomposition method; calculating and estimating instantaneous fluctuation characteristics of a fundamental frequency of the extracted rotating frequency component using a quadrature-derivative-based normalized Hilbert transform to obtain an instantaneous frequency; inputting a calculated instantaneous frequency sequence into an optimum stochastic resonance system for anti-noise enhancement of intrawave frequency modulation characteristics; and processing FFT on an instantaneous frequency of intrawave frequency modulation characteristics after anti-noise enhancement, and diagnosing and identifying a rotor rub-impact fault of the rotating machinery based on a distribution characteristic of harmonic amplitudes related to a rotor rotating frequency.

The invention relates to the field of artificial intelligence hardware acceleration, and discloses a memristor-based multilevel compression collaborative optimization neural network deployment method and device, and a storage medium. The method comprises the following steps: carrying out progressive structured pruning on a pre-training model based on a dual-drive scoring mechanism of an L1 norm and gradient sensitivity and hardware feedback, and generating a hardware-friendly sparse weight structure; the characteristics of the memristor are simulated through a micro-nonlinear conductance modeling function, and network weight and conductance parameters are synchronously optimized to reduce errors by adopting mixed precision quantification of four bits of a convolutional layer and two bits of a full-connection layer; a conductance drift and read-write noise model is injected, and the anti-interference capability of the model is improved in combination with adaptive noise enhancement and KL divergence loss; and mapping the optimized model to a memristor memory architecture to complete weight coding and reasoning. Through collaborative optimization of pruning, quantification and distillation, the problems of insufficient storage density, non-ideal characteristic interference and algorithm and hardware mismatch in memristor deployment are solved.

The invention discloses a multi-unmanned aerial vehicle formation dynamic obstacle avoidance method and device and a storage medium, and belongs to the field of unmanned aerial vehicles. According to the method, the dynamic obstacle speed relevance is modeled by adopting the space-time interaction graph, and the graph convolutional network and a noise enhancement mechanism are combined, so that high-precision trajectory prediction is realized. And designing a dual-channel generative adversarial network, synchronously optimizing the path smoothness and the space-time conflict risk, and evaluating the security of the multi-machine collaborative path in real time in combination with an attention network. The formation control layer breaks through the limitation of traditional local optimization, establishes a multi-agent game model, dynamically balances obstacle avoidance, formation keeping and energy consumption through a utility function, and realizes formation form stability in combination with a centralized training framework. And aiming at a severe environment perception deviation problem, a cross-domain feature alignment technology is adopted, so that the recognition robustness in scenes such as rain and fog is effectively improved. A complete technical closed loop from environmental perception, dynamic prediction to collaborative decision is constructed, and systematic theoretical support is provided for autonomous obstacle avoidance of an unmanned aerial vehicle cluster in a dense dynamic scene.

The invention relates to the field of knowledge maps, and discloses a multi-modal entity alignment method and device and a medium, and the method comprises the steps: obtaining the data of two multi-modal knowledge maps, preprocessing the corresponding data, and obtaining the preprocessed data; constructing a multi-modal encoder containing a modal adaptive noise enhancement mechanism, enhancing the expression ability of multi-modal data through Gaussian noise injection, and inputting the preprocessed data to the multi-modal encoder to obtain enhanced multi-modal embedding features; calculating a single-modal confidence coefficient and a joint confidence coefficient for the enhanced multi-modal embedded feature to obtain a modal weight; through a relative calibration strategy, the weight of a fusion mode is adjusted according to the uncertainty of the mode, and the upper bound of a generalization error is reduced; multi-modal joint embedding is obtained, and entity alignment is carried out by combining intra-modal and inter-modal comparison loss; according to the method and the device, the technical problem of wrong alignment caused by modal quality difference, noise and inconsistency and isomerism among modals in a multi-modal entity alignment method is solved.

The invention discloses a noise-driven small sample distribution external detection method and system, and the method comprises the following steps: 1, obtaining a data set which comprises a small sample training set and a test set; 2, constructing a detection model, wherein the detection model comprises a data coding module, a noise extraction module and a fusion detection module; 3, performing noise enhancement on the noise region set, constructing a loss function in combination with the clean region set, training the detection model by adopting a gradient back propagation algorithm according to the loss function, and testing the detection model by adopting the test set; 4, performing small sample distribution external detection according to the tested detection model; according to the method, the problem of poor noise robustness of a current small sample distribution external detection method is solved, and the ID classification and OOD detection performance of a small sample learning model is improved.

The invention discloses a privacy data protection method and device based on differential privacy, equipment and a medium. The intelligent contract containing the differential privacy strategy is integrated to an intelligent contract layer of the block chain, the differential privacy strategy is used for calling the intelligent contract through the intelligent contract layer to automatically identify sensitive data in the uploaded financial data when the financial data is uploaded, and noise adding processing is performed on the sensitive data; storing the sensitive data subjected to noise addition processing in a public account book for all nodes of the block chain to access; original financial data and noise parameters generated in the noise adding process are encrypted and then stored in a private account book, and the private account book can only be accessed by an authorized node; and creating at least one role and the access authority corresponding to each role to perform authority management on the access of the private account book. According to the scheme, secure sharing and efficient use of financial privacy data can be realized.

The invention relates to the technical field of image processing, and provides an underwater image enhancement method based on frequency domain enhancement and edge guidance, which comprises the following steps of: firstly, acquiring three images with different scales through image downsampling, and then converting the three images into a frequency domain to respectively extract a low-frequency spectrum and a high-frequency spectrum; constructing a distance mask and a mean value mask based on the distance and the amplitude mean value to enhance a high-frequency part so as to improve image details; meanwhile, the effect is enhanced by suppressing low-frequency part noise. In addition, a multi-stage residual feature aggregation module is provided, the module focuses on detail extraction, and information loss caused by global enhancement is effectively avoided. And finally, further enhancing image edge details in combination with an edge guiding strategy. Experimental results show that the method is superior to the current most advanced underwater image enhancement method in quantitative and qualitative evaluation of a plurality of public data sets.

The invention provides a far and near field multi-axis sensing array coupled converter transformer monitoring method and system. The system comprises a near field four-axis sensing array, a far field four-axis sensing array, a signal-to-noise improving unit, an original point processing unit, an AD conversion unit and a background display control seat unit. Acquiring near-field and far-field noise signals without excitation, and coupling the near-field noise signals; taking the far-field noise signal as input, taking the coupled near-field noise signal as output, and training a deep network model; acquiring a to-be-detected signal and a field noise signal under excitation, and coupling the to-be-detected signal; taking the field noise signal as the input of the trained deep network model to obtain a reconstructed coupling noise signal; the method comprises the following steps: obtaining a denoised signal to be detected and a denoised coupling signal by using a coupling noise signal, carrying out signal processing to obtain a corresponding analog signal, converting the analog signal into a corresponding digital signal, and carrying out visual monitoring. The fault recognition rate and the operation and maintenance efficiency are remarkably improved.

The invention provides a privacy computing service method and system based on federal learning. According to the method, desensitized data is generated by identifying sensitive field types of original data of federated learning participants, dividing privacy levels and carrying out local shielding and superposition of dynamic noise. Dividing the data into data blocks in direct proportion to the number of the arithmetic units based on the complexity of an encryption algorithm, performing parallel encryption, transmitting the data blocks to the cooperative node through an encryption channel to obtain an aggregation result, and synchronously recording time consumption and a privacy intensity value. The data block segmentation number is dynamically adjusted by calculating the time consumption difference value, and correlation parameters are generated through cross validation with the privacy intensity value. When the privacy intensity is insufficient, the dynamic noise of the corresponding level is improved, data block segmentation is synchronously optimized, and a dynamic linkage mechanism of noise enhancement and block encryption is formed. Through a dynamic mutual feedback mechanism of the difference value and the privacy intensity, real-time balance between the data desensitization intensity and the encryption calculation efficiency is realized.

The invention belongs to the technical field of multimedia analysis, and discloses a multi-modal media tampering detection method, system and device based on multi-view comparative learning and a medium, and the method comprises the steps: obtaining a training data set which comprises a training image-training text pair and a corresponding tampering category label; a cross encoder is introduced on the basis of a vision-language model, a plurality of multi-layer sensor head structures are arranged, and three kinds of comparative learning of noise enhancement, prototype-based and multi-label tampering classification are designed to obtain an initial multi-view comparative learning framework; training the initial multi-view comparative learning framework based on the training data set to obtain a trained multi-view comparative learning framework; and based on the trained multi-view contrast learning framework, executing a tampering detection task of the to-be-detected image-text to the data. According to the technical scheme, the accuracy and robustness of multi-label classification can be improved.

This invention discloses a bioacoustic complexity index method with low sensitivity to noise, comprising: performing a short-time discrete Fourier transform on the field bioacoustic monitoring recording data during the analysis period to obtain its time-frequency power spectrum; estimating the narrowband noise average power of each frequency point using a set of noise time-frequency points without bioacoustic signals; performing power spectrum subtraction on each frequency point to obtain the noise-enhanced bioacoustic time-frequency power spectrum; subdividing the analysis bandwidth and analysis period into time-frequency domains according to a certain frequency step and time step, and calculating the time-frequency partition index value; and summing all the time-frequency partition index values ​​within the analysis bandwidth and analysis period to obtain the final bioacoustic index value, i.e., the noise-reduced bioacoustic complexity index based on the change in bioacoustic time-frequency power after noise enhancement. This invention significantly reduces the sensitivity of the index results to noise and significantly expands the spatiotemporal application range of similar acoustic index strategies.

The invention discloses an oil-immersed current transformer data enhancement method based on transfer learning. The method comprises the following steps: constructing a data preprocessing model, and carrying out standardization processing on oil chromatographic data of a transformer and a current transformer by utilizing a quantile transformation method; constructing an encoder to map original signal data into a potential variable sample space, and proposing a noise enhancement mechanism with multi-distribution fusion; the generative network performs random sampling from known probability distribution to obtain hidden variables, and calculates a preliminary enhanced signal obeying Gaussian distribution; the generative adversarial module adopts a generator and a discriminator, the generator performs incremental generation on the preliminary enhanced signal and outputs a sample conforming to the characteristics of the current transformer, and the discriminator distinguishes the difference between a real current transformer sample and a generated sample according to physical knowledge constraints in the oil chromatography field; and finally, carrying out reverse quantization on the generated sample, and outputting high-quality oil chromatographic data of the current transformer. According to the invention, the problem of rare data of the current transformer is effectively solved.

The invention provides a monitoring method based on blade rotation sound and equipment for auxiliary monitoring, and the method comprises the steps: capturing weak airflow disturbance sound wave signals generated during the rotation of blades of a wind driven generator through a non-contact acoustic sensor array, and carrying out the standardization preprocessing of the captured sound wave signals; the method comprises the following steps: performing dynamic filtering processing on environmental noise of sound wave signals, integrating acoustic sensor data from different positions to form a comprehensive sound image, applying deep learning to model and predict the health condition of a blade, and integrating edge calculation, so that a part of data processing tasks are completed in a place close to an acoustic sensor; the Internet of Things platform is adopted to realize interconnection and intercommunication between equipment, implement a cloud edge collaborative optimization mode and automatically allocate tasks to the most appropriate processing node, sensors do not need to be installed on the blades, physical influences on the equipment are avoided, environmental noise is effectively filtered out, the signal-to-noise ratio of target sound wave signals is increased, and therefore the monitoring precision is improved.

The invention relates to the technical field of transformer fault diagnosis, and particularly provides a transformer fault diagnosis method based on a golden vehicle optimization algorithm. The method comprises the steps that a GJO-SDAE joint optimization network model is constructed, classification accuracy and reconstruction errors serve as fitness functions, and network structure parameters and hyper-parameter combinations of SDAE are synchronously optimized; designing an anti-noise enhancement mechanism through a GJO-SDAE combined optimization network model, and meanwhile, designing a weighted fusion strategy of multi-level noise reduction features; according to the method, the optimized network model is obtained through GJO, meanwhile, an interpretability module based on feature importance is introduced, a diagnosis decision basis with semantic tags is generated by analyzing a weight matrix optimized through GJO, and the transformer fault diagnosis performance is improved through the method.

The invention discloses a U-Net seismic data denoising method based on multi-scale fusion boundary feature enhancement, and the method specifically comprises the steps: 1, preliminarily constructing a seismic data training data set A which is not subjected to noise enhancement, and carrying out the preprocessing of the preliminarily obtained seismic data training data set A which is not subjected to noise enhancement, obtaining a seismic data training data set B which is finally subjected to noise enhancement; step 2, building a U-Net network framework with enhanced multi-scale fusion boundary features; step 3, performing parameter configuration and optimization setting on the U-Net network with enhanced multi-scale fusion boundary features constructed in the step 2; and step 4, training and optimizing the U-Net network enhanced by multi-scale fusion boundary features. The method can retain more weak signals and detail information while suppressing seismic noise.

The invention discloses a speech recognition method and device based on environmental noise enhancement and a medium, and the method comprises the steps: carrying out the framing processing of a to-be-recognized target speech signal, and obtaining at least one frame of speech signal; determining respective first power spectrums of the at least one frame of voice signal, and further determining respective Mel spectrums of the at least one frame of voice signal; determining respective noise feature vectors of the at least one frame of voice signal based on a noise perception model; and target model parameters of the speech enhancement model are determined based on the parameter generation model, so that the model parameters of the speech enhancement model are adaptively generated according to the to-be-recognized target speech signal. Then, inputting the respective first power spectrum of the at least one frame of voice signal into a voice enhancement model, and determining the respective second power spectrum of the at least one frame of voice signal based on the voice enhancement model; and finally, determining a target text sequence corresponding to the target voice signal based on the voice recognition model. Therefore, the accuracy of the determined target text sequence is improved.

The invention discloses a noise enhancement nonlinear system joint detection and estimation method under a Bayesian framework, and belongs to the field of signal processing. Firstly, independent additive noise is added to a nonlinear system input signal, and noise-corrected nonlinear system output is obtained after the signal passes through a nonlinear system. And secondly, under the Bayesian criterion, judging which hypothesis in the binary hypotheses is established by utilizing the output of a nonlinear system of noise correction, and estimating unknown parameters in the signal of which the judgment result is H1. On the premise that the detection performance is not reduced, a noise enhancement nonlinear system joint detection and estimation model which minimizes the estimation risk is constructed. The additive noise is the optimal solution of the model and is random distribution formed by not more than two constant vectors. According to the method, noise enhancement and nonlinear system joint detection and estimation under the Bayesian framework are combined, and the Bayesian estimation risk is further reduced under the condition that the Bayesian detection cost is not increased.

An information processing apparatus including: a constraint loss calculation unit that calculates a constraint loss using an estimated speech enhancement mask output by a speech enhancement mask estimation model in case a noise-mixed speech is input, and an estimated noise enhancement mask output by a noise enhancement mask estimation model in case the noise-mixed speech is input; and a parameter update unit that updates parameters included in the speech enhancement mask estimation model and parameters included in the noise enhancement mask estimation model, using a speech enhancement mask loss indicating a difference between the estimated speech enhancement mask and a target speech enhancement mask calculated in case a speech and a noise in the noise-mixed speech are known, and a noise enhancement mask loss indicating a difference between the estimated noise enhancement mask and a target noise enhancement mask calculated in case a speech and a noise in the noise-mixed speech are known, and the constraint loss.

The invention discloses a noise monitoring method and system based on voiceprint recognition, and relates to the technical field of noise monitoring. Acquiring all uploaded initial noise data and coordinate positions to obtain a noise environment data set; performing feature extraction on the uploaded initial noise data to obtain logarithm Mel spectrum features and time-frequency features; substituting the logarithmic Mel spectrum features and the time-frequency features of all the effective noise acquisition nodes into a voiceprint recognition model to obtain a noise category set; performing noise enhancement on the noise categories with the same frequency labels to obtain enhanced noise features, and combining the enhanced noise features with each noise environment data group to obtain a noise positioning group; and substituting the noise positioning group marked with the same noise category into the noise positioning model to obtain a noise position. According to the method, invalid background data are removed, then logarithmic Mel spectrum and time-frequency characteristics are extracted, similar noise positioning data are substituted into a model in a concentrated manner, data cross interference is avoided, and finally noise identification and positioning accuracy is improved and monitoring efficiency is optimized.

This invention belongs to the field of multimedia analysis technology and discloses a multimodal media tampering detection method, system, device, and medium based on multi-view contrastive learning. The method includes: acquiring a training dataset, which includes training image-training text pairs and corresponding tampering category labels; introducing a cross-encoder based on a visual-language model, setting several multilayer perceptron head structures, and designing three contrastive learning methods: noise enhancement, prototype-based, and multi-label tampering classification, to obtain an initial multi-view contrastive learning framework; training the initial multi-view contrastive learning framework based on the training dataset to obtain a trained multi-view contrastive learning framework; and performing a tampering detection task on the image-text pair data to be detected based on the trained multi-view contrastive learning framework. The technical solution of this invention can improve the accuracy and robustness of multi-label classification.

This application belongs to the field of interconnection network reliability and fault diagnosis technology, and discloses a method for intermittent fault diagnosis of low-bandwidth long interconnection networks based on graph attention mechanism. Targeting the hierarchical recursive structure and high connectivity of the network, under the PMC fault diagnosis model, a multi-round testing strategy is used to obtain test symptoms within the node's neighborhood. A feature vector is constructed for each node using local statistical feature extraction methods, and preprocessed using zero-padding and noise enhancement techniques. Finally, a graph attention network model is constructed, and the importance weights of neighboring node test results are dynamically learned using the attention mechanism to achieve accurate diagnosis of node fault states in the network. This application leverages the powerful local information aggregation capability of graph attention networks to overcome the diagnostic limitations of traditional algorithms, maintaining high diagnostic accuracy and robustness even with high fault rates, incomplete test symptoms, and large network sizes.

The invention discloses a remote sensing image unsupervised change detection method and system based on structure perception and noise enhancement. According to the method, shared feature representation is extracted from a dual-temporal remote sensing image, a structure perception contrast learning mechanism is introduced to enhance the perception capability of a model for real geographic structure change, noise disturbance consistency constraint is designed to avoid an optimization shortcut problem, and a frequency attention decoding mechanism is adopted to finely depict a change region boundary. The system comprises a preprocessing module, a feature coding module, a structure sensing module, a noise disturbance module, a frequency attention decoding module and an unsupervised optimization module. According to the method, under the condition that manual labeling is not needed, the problems that an existing unsupervised change detection method is short in optimization, insufficient in semantic representation capacity, not fine in boundary description and the like are effectively solved, the accuracy and robustness of change detection are improved, and the method is suitable for the fields of urban expansion monitoring, disaster assessment, environment change analysis and the like.

The invention aims to provide a voice training noise adding system and method based on a mixed noise generation model. The system comprises an input module, a noise environment enhancement module, a voice noise enhancement module and an output module. Wherein the input module is used for acquiring noise environment simple description information and clean voice data to be enhanced; the noise environment enhancement module converts the noise environment simple description information into a structured noise event sequence with time sequence characteristics; a voice noise enhancement module generates multi-source mixed noise according to the noise event sequence, and adds the multi-source mixed noise to the clean voice data according to a preset rule to obtain noisy voice data; and the output module is used for outputting the noisy voice data for voice model training. According to the invention, interaction characteristics of superposition, offset, interference and the like of different noise sources in the time dimension are fully joined, and the technical bottleneck that only linear superposition can be realized in a traditional mixing mode is solved.

The application discloses a ship radiation noise enhancement method and system, electronic equipment and a storage medium, and relates to the technical field of radiation noise enhancement. The array receiving signal is sequentially subjected to time delay beam forming and time sliding window segmentation to obtain a plurality of time direction signal vectors, and a plurality of to-be-processed signal vectors are selected from the time direction signal vectors. For a FrFT domain real signal determined after processing of any to-be-processed signal vector by using a plurality of different fractional Fourier transform algorithms, fitting goodness tests of Alpha stable distribution and asymmetric Laplace distribution are respectively performed to obtain a first D value and a second D value, and a target order and a target statistical distribution are determined. Any time direction signal vector is processed by using a fractional Fourier transform algorithm with the target order to determine a target FrFT domain real signal. According to the target statistical distribution and all target FrFT domain real signals, a time bearing history graph of the array receiving signal is determined, and the enhancement of the ship radiation noise is realized.

The invention relates to the field of noise reduction of centrifugal pumps, and discloses an internal noise reduction type centrifugal pump which comprises a pump cover, the volute is arranged on one side of the pump cover, and an impeller is arranged in the volute; the noise reduction box body is arranged in the pump cover, and a motor is arranged in the noise reduction box body; through multiple noise reduction and vibration reduction structures, the noise reduction capacity of the centrifugal pump is improved, and the service life of the centrifugal pump is prolonged. The motor is packaged through the noise reduction box body, a first sound insulation barrier is formed, outward transmission of noise is reduced, meanwhile, the sound insulation cotton a is arranged in the noise reduction box body to wrap the motor, and high-frequency noise generated when the motor works can be effectively absorbed. In addition, a sound insulation cavity is formed in the pump cover and filled with sound insulation cotton b, so that propagation of low-frequency noise is reduced, and residual noise is further attenuated. The cushioning component is arranged between the motor and the noise reduction box body and can absorb mechanical vibration during operation of the motor and prevent the vibration from being directly transmitted to the noise reduction box body to cause a resonance amplification effect, so that noise enhancement and structural fatigue caused by resonance are reduced.

The invention provides an abnormal sound sample generation method, and a part abnormal sound fault diagnosis method, device and equipment, and relates to the technical field of abnormal sound fault diagnosis. And generating target noise according to preset parameters, fusing the target noise with the real abnormal sound audio data spectrogram, and generating noisy abnormal sound spectrograms under different noise interferences through a gradual noise adding diffusion process. Sequentially executing N times of feature coding, channel dimension self-adaptive calibration and spatial dimension self-adaptive calibration on the noisy abnormal sound spectrogram, generating N joint calibration feature maps fusing noise suppression and abnormal sound enhancement layer by layer, and after feature coding, executing N times of up-sampling operation in combination with the joint calibration feature maps generated at the corresponding levels, and generating an abnormal sound sample set which is highly close to the real abnormal sound audio data. The problems of weak generalization ability of a diagnosis model and low detection accuracy in diversified actual working conditions caused by insufficient real abnormal sound fault samples are solved, and the accuracy and stability of abnormal sound fault diagnosis of parts are enhanced.

The invention discloses a digital twinning edge enabling smart factory task unloading method based on a near-end policy optimization (PPO) algorithm, and relates to the technical field of computers, and the method comprises the steps: abstracting a current smart factory scene into a digital twinning edge enabling smart factory network model, and carrying out the task unloading of the digital twinning edge enabling smart factory network model; modeling tasks in the smart factory to obtain a multi-constraint task model; establishing a multi-constraint task unloading problem model including time delay, energy consumption and task completion rate according to the power of the local equipment, the channel transmission bandwidth, the computing power of the server and the fault condition; a PPO network is designed based on a multi-constraint task model and a multi-constraint task unloading problem model, and a smart factory task unloading problem is processed into a Markov decision process including a state, an action and a reward; the method comprises the following steps of: establishing a PPO (Point-to-Point Optimization) model, adding uniform noise into the PPO network, establishing a Noise-Enhanced PPO (Point-to-Point Optimization) algorithm for noise enhancement according to the model, and solving to obtain an optimal task unloading decision.

The invention discloses a compound fault diagnosis system and method for a centrifugal blower, and belongs to the field of signal fault diagnosis of rotating machinery. The method comprises the following steps: collecting a vibration signal through a multi-position acceleration sensor; decomposing the signal by using empirical mode decomposition, and preliminarily screening IMF components; performing noise reduction by using an arc tangent threshold function; carrying out fault feature enhancement by using an optimized bistable stochastic resonance system; performing feature level fusion on the envelope spectrum of the signal after noise reduction processing and the frequency spectrum of the signal after feature enhancement in a frequency domain to obtain a fusion feature; and inputting the fusion features into a width learning system to realize compound fault diagnosis of the centrifugal blower. According to the method, the self-adaptive decomposition capability of empirical mode decomposition, the noise enhancement characteristic of stochastic resonance, the characteristic complementarity of frequency domain fusion and the rapid classification advantage of a width learning system are integrated, and the centrifugal blower composite fault diagnosis work under the complex working condition can be better completed.

The application provides a kind of electrically conductive sponge electromagnetic leakage dynamic monitoring method and system, by utilizing the leakage feature recognition degree of auxiliary noise enhancement based on spectrum attenuation coefficient, and utilizing the set threshold value, a plurality of IMF components are obtained by ensemble empirical mode decomposition, for each IMF component, determine the dominant frequency point and phase retention band;While the phase is retained in the retention band, the out-of-band phase is randomized, the proxy data set is constructed and the confidence interval is established, so as to screen out the leakage feature component, by extracting the weighted hilbert energy with the energy proportion in the phase retention band as the weight, and the multiscale sample entropy with the total bandwidth as the scale factor, a joint feature vector is constructed, and the leakage is judged based on the mahalanobis distance of the vector and the normal state.

The application provides a spoken English recognition method and system based on contrast learning and mixed attention, which comprises the following steps: acquiring audio of a spoken English test, adding noise of a random category to environmental recording to realize noise enhancement and further construct a positive sample; performing feature extraction on the data added with noise based on multi-scale and mixed attention; inputting the features after embedding and position coding into an encoder to perform context modeling; inputting the output of the encoder and the target features after embedding and position coding into a decoder to complete decoding; in the training process, calculating a contrast loss through input of the positive sample, calculating the loss of each sample at the same time, performing reverse transmission, and obtaining a recognition model; and inputting the audio of a candidate to be transcribed into the recognition model to obtain a recognition result.