178 results about "Noise estimation" patented technology

According to the VLA-based body robot SLAM method and device and the storage medium, a VLA large model is introduced on the basis of multi-modal fusion SLAM of traditional point cloud geometry, vision and the like, perception is improved from a geometric layer to semantic concept alignment, and the SLAM is more accurate. A VLA large model is used for carrying out dynamic prediction updating on dynamic interference filtering, key frame screening, factor graph relation construction, noise estimation, loopback detection and the like, the real-time requirement is met in the modes of incremental optimization and the like, and the dislocation problem of geometric constraints is corrected through global semantic constraints. The stability of the robot SLAM in extreme scenes such as excessive environmental dynamic interference, loud sensor noise and environmental degradation is improved, and the constructed hierarchical situation map can meet the requirement of a high-order navigation task while the geometric accuracy is met, so that the robot SLAM can be deployed to carriers such as a body-equipped intelligent carrier for subsequent application.

The invention provides an inertia / satellite / vision adaptive integrated navigation method and system. The method comprises the following steps: carrying out inertia measurement and navigation calculation to obtain inertia data; calculating satellite navigation observed quantity according to the satellite navigation data and the inertial data, and constructing a satellite observation model; calculating visual navigation observed quantity according to the visual navigation data and the inertial data, and constructing a visual observation model; aiming at a satellite and a visual observation model, respectively adopting an innovation-based adaptive covariance estimation method to obtain satellite and visual observation noise covariance updated values; constructing a satellite and visual navigation health degree, and calculating a satellite and visual fusion weight according to the satellite and visual navigation health degree; controlling updating of satellite and visual navigation observed quantity according to the satellite and visual fusion weight; and calculating a weighted equivalent observation matrix and a noise covariance, and carrying out filtering estimation. According to the method, an adaptive noise estimation and sensor health degree evaluation mechanism is introduced, dynamic weighted fusion of multi-sensor data is realized, and the robustness and precision of a navigation system in complex environments such as satellite signal lock losing and visual feature missing are improved.

The invention discloses a semantic constraint adversarial sample generation method and system. According to the method, firstly, a confrontation sample generation task based on a natural language instruction is constructed, and parameters are initialized; then deploying a proxy model and a diffusion model for double-branch noise estimation; by judging instruction complexity, mask guidance is selectively adopted to realize natural constraints of spatial differentiation; by constructing a residual-guided antagonistic DDIM sampler and combining an adaptive optimization iterative algorithm, the calculation complexity is reduced, and the attack mobility is enhanced at the same time; and aiming at a three-dimensional generation requirement, a three-dimensional Gaussian sputtering rendering model is further integrated, and geometric consistency is ensured through multi-view gradient averaging. According to the method, the problems of inaccurate semantic control, weak migration aggressiveness, poor three-dimensional generation consistency and the like in the prior art are effectively solved, the attack success rate and the visual naturalness are remarkably improved on multiple target models, and an efficient and reliable red team test tool is provided for security evaluation and alignment of a multi-modal large model.

The invention relates to a fixed parameter optimization cubature Kalman filtering tracking method for a high-speed high-maneuvering target, belongs to the technical field of signal processing and target tracking, and aims to solve the problem of insufficient tracking performance caused by difficulty in parameter tuning and isolation of an improved mechanism when an existing method is used for coexistence of model mismatch, noise time variation and outlier interference. According to the scheme, a framework integrating off-line multi-parameter collaborative optimization and on-line multi-mechanism adaptive filtering is constructed; in the off-line stage, an optimal combination of key parameters such as covariance adjustment factors is determined through a grid search system; in the online stage, the combination is loaded, and a complete filtering process including innovation feedback type dynamic covariance adjustment, sliding window type noise estimation and outlier suppression and trace-related adaptive regularization is executed, so that an enhanced tracking method with active pre-judgment and closed-loop learning capabilities is formed. The method is mainly used for carrying out high-precision and high-robustness real-time state estimation on the high-speed high-maneuvering target.

The invention provides a high-throughput Raman spectrum-oriented identification method, and belongs to the technical field of Raman spectrum data processing. According to the method, the number of principal components is dynamically determined through the adaptive feature contribution rate and the noise estimation algorithm, flexible compression for different spectral complexity is achieved, and the operation efficiency in large-batch data processing is remarkably improved; on this basis, a lightweight convolutional neural network model introducing a channel attention mechanism is constructed to enhance the response ability to a key Raman peak position and suppress background interference; and meanwhile, an auxiliary loss function based on inter-spectrum cosine similarity constraint is introduced in the training process, the intra-class consistency and the inter-class separation degree are improved, and a visual feedback mechanism is combined to optimize the region of interest of the model on a discrimination section. Compared with a traditional Raman spectrum classification method, the method does not depend on artificial feature extraction and preset dimension parameters, has the advantages of being high in discrimination, good in interpretation, wide in adaptability and the like, and can achieve efficient and accurate classification of complex Raman spectrums.

The invention relates to a dynamic light beam shaping system and method based on a spatial light modulator. The system comprises a laser light source module, a beam expanding collimation light path, a light beam modulator, a focusing optical assembly, a detection and feedback module and a control and optimization module. According to the method, the uniformity of the output flat-topped beam can be corrected by virtue of a random parallel gradient descent (SPGD) algorithm and combining a Zernike polynomial as a control variable. Meanwhile, a composite performance index comprehensively considering energy distribution and shape features is constructed, the correction process of the flat-topped beam is accelerated by adopting a mode of gradually improving Zernike polynomial orders in multiple stages, the convergence trend of the performance index is used as a judgment basis of stage switching, stage independent variable dimension expansion is implemented, and the correction precision of the flat-topped beam is improved. And the convergence speed is obviously improved. Besides, adaptive selection of disturbance amplitude and learning rate parameters in the SPGD algorithm is realized by using noise estimation and a local linearity index, and the robustness of the system to noise and environmental disturbance is enhanced.

The invention discloses a joint blind denoising method and system based on self-heuristic learning and Bayesian reasoning, belongs to the field of computational imaging, and solves the problems that in the prior art, the mixed noise modeling capability is insufficient, the performance is degraded under the condition of low signal-to-noise ratio, the combination of uncertainty quantization and regularization is lacked, and the generalization capability is limited due to data dependence. Comprising the following steps: collecting an original image and preprocessing; generating a noise data pair; an enhanced residual attention U-Net model is constructed; a noise estimation sub-network is adopted to extract noise features, the noise features are fused with original image features, and the model is trained; adopting the trained model to carry out multiple times of forward propagation on the same input image to obtain multiple groups of denoising results; calculating a mean value and a standard deviation to obtain a de-noising prediction and uncertainty heat map; and training the trained model again based on the uncertainty heat map, optimizing network parameters, and obtaining a final denoising prediction result and uncertainty estimation thereof. The method is suitable for complex noise distribution processing scenes.

The invention discloses a sound signal processing method and device, equipment and a storage medium, and belongs to the technical field of audio processing. According to the invention, sound signal noise reduction with noise suppression and target signal reservation is realized. The method comprises the following steps: after acquiring a sound signal collected in a running state of mechanical equipment, firstly performing time-frequency analysis on the sound signal; then, a noise determination threshold is automatically determined based on the logarithmic magnitude spectrum of the sound signal, and noise estimation is performed based on the determined noise determination threshold. According to the scheme, a completely data-driven parameter selection mechanism is realized, and manual parameter or threshold setting is not needed, so that the automation degree is improved, the unreliability of manual parameter or threshold setting is avoided, and the accuracy and robustness of noise estimation are enhanced. In addition, the noise-reduced sound signal does not comprise noise components, so that the accuracy and reliability of subsequent operation state recognition and fault diagnosis of the mechanical equipment are ensured.

The invention relates to the technical field of acoustic signal processing, and discloses a sound source positioning method and system, and the method comprises the steps: collecting sound through a multi-microphone array, aligning data based on a synchronization mechanism, and obtaining an initial time domain signal; executing transformation and filtering to generate a spectrogram; separating interference based on noise estimation, and extracting independent source frequency domain components; calculating a time difference in combination with a synchronization mechanism, and determining a preliminary orientation; the moving direction is analyzed, the track is smoothed, and continuous track data are obtained; constructing a velocity vector to generate a Doppler compensation coefficient, and executing phase calibration to obtain a calibration signal; and performing inverse transformation on the reconstructed signal, and calculating a generalized cross-correlation peak value to determine a final position. The method can achieve the precise tracking and positioning of a dynamic sound source, effectively eliminates the Doppler effect and environment noise interference, and remarkably improves the positioning precision.

The invention discloses a magnetocardiogram P-wave detection method based on wavelet analysis and a local noise self-adaptive threshold, which is characterized in that accurate and robust detection of a P wave is realized by searching on a characteristic scale which can highlight the P wave, a self-adaptive threshold confirmation mechanism based on local noise estimation is introduced, and a self-adaptive threshold confirmation mechanism is established for a time domain search window of each cardiac cycle. According to the method, the internal noise level is dynamically evaluated, and an estimated value capable of accurately reflecting the local noise condition of the current heart beat can be obtained by carrying out statistical analysis on wavelet coefficients in a characteristic scale in an area where P waves are most likely to appear.

The invention discloses a molded case circuit breaker fault prediction and alarm method based on artificial intelligence, and the method comprises the following steps: S1, collecting current, voltage, temperature and contact resistance signals, constructing multi-channel time sequence observation data, and generating a fusion observation vector; s2, inputting the fusion observation vector to a noise estimation sub-network, and generating a dynamic covariance parameter; s3, inputting the fusion observation vector, the dynamic covariance parameter and a previous state estimation value into a KalmanNet structure, and outputting a current state estimation value; s4, a physical prior regularization module is introduced in the training stage, a constraint loss function is constructed, and network parameters are jointly optimized; s5, executing drift detection in an operation stage, and extracting historical window data to perform incremental updating when conditions are met; s6, pruning and quantifying the trained KalmanNet structure, and generating a lightweight model; and S7, deploying to a monitoring system, and predicting the state in real time for alarm judgment. According to the invention, the accuracy and deployment efficiency of circuit breaker fault prediction are improved.

The invention provides a single photon array image data processing method and system based on time correlation, and relates to the technical field of image processing.The method comprises the steps that a photon event sequence with a timestamp and position information is obtained through a detector array, a four-dimensional space-time tensor is constructed, and space-time data is obtained through normalization and noise estimation; performing time sequence analysis on each spatial position and neighborhood in the spatio-temporal data to extract time sequence features, separating target features from the time sequence features, removing noise features and generating a time confidence map; inputting the time confidence map and the spatio-temporal data into a variational optimization model for joint optimization, and dynamically updating a time weight field and a spatial support domain in optimization according to the confidence map to obtain an initial reconstructed image; and finally, inputting the initial reconstructed image and the time confidence map into a neural network, and outputting a final image through modeling and residual path restoration. According to the invention, the signal-to-noise ratio and dynamic scene adaptability of single photon imaging under extremely low illumination are improved.

Transformer neural network based decoder for decoding Quantum Error Correction Codes (QECC), comprising, an input layer, a plurality of decoding layers, and an output layer. The input layer is adapted to receive initial noise estimation computed by a noise estimator for noise injected to syndrome bits of codewords encoded using QECC and transmitted over transmission channel(s) subject to interference, and create embeddings for the syndrome bits. The decoding layers adapted to compute an estimated logical operator matrix of each codeword, each comprises a self-attention layer constructed according to a mask indicative of a relation between the embeddings derived from a parity-check matrix of the error correction code. The plurality of decoding layers are trained using a combined loss function directed to minimize LER, BER, and error rate of the noise estimator. The output layer is adapted to produce a vector representing predicted soft error of the codeword's logical operator matrix.

The invention discloses a Bluetooth earphone call intelligent noise reduction method based on cloud cooperation, and the method comprises the following steps: S1, obtaining a call voice signal and an environment state parameter, carrying out the preprocessing, and constructing a voice spectrum sequence and an environment feature sequence; s2, performing noise estimation and suppression processing on the speech spectrum sequence by using a DCCRN model to obtain a spectrum feature sequence; s3, constructing a cooperative processing sequence based on the environment feature sequence and the spectrum feature sequence; s4, using a Conv-TasNet model to execute speech enhancement on the co-processing sequence, and generating a target spectrum sequence; s5, performing phase reconstruction and inverse frequency spectrum transformation based on the target frequency spectrum sequence to generate a target voice signal; and S6, constructing a loss function according to the target voice signal, and updating the Conv-TasNet model. According to the method, the DCCRN model and the like are fused, and the method has the advantages of high noise reduction precision, high environment adaptability and sustainable optimization.

The invention relates to a spatio-temporal data rule extraction method based on spatio-temporal Fourier expert mixing, belongs to the technical field of urban spatio-temporal data processing, solves the problem that global stability characteristics and local disturbance characteristics driven by real physical rules are difficult to effectively describe in the prior art, and comprises the steps that S1, a spatio-temporal data acquisition module acquires historical data information; s2, establishing a space-time Fourier expert hybrid network, and carrying out expert mixing to obtain mixed features; s3, using adaptive group normalization as a conditional fusion module to obtain fusion features; s4, establishing a space-time Fourier attention mechanism module to obtain output features; s5, establishing a diffusion model for performing back diffusion on the noisy data in combination with the noise estimation network to obtain predicted denoised data, and performing training to obtain a trained diffusion model; and S6, performing sampling to obtain spatio-temporal data, inputting the spatio-temporal data into the trained diffusion model, and obtaining an extracted causal law for urban traffic flow prediction.

This invention belongs to the field of medical image processing and deep learning inference technology, specifically relating to a method, device, and storage medium for denoising and enhancing low-dose CT images. The denoising and enhancement method includes the following steps: S1, acquiring a low-dose CT image; S2, calculating the noise level parameter N; S2.1, generating a uniform reference region mask; S2.2, robust noise estimation within the mask; S2.3, robust fusion of the noise level parameter; S3, generating a denoised image I_dn based on the CT image to be processed using a deep learning denoising network; S4, calculating the denoising intensity coefficient s according to the noise level parameter N, and performing weighted fusion of the CT image to be processed and the denoised image I_dn to obtain the output enhanced image. This invention has the advantages of strong adaptability, no need for manual ROI selection, good robustness, and easy deployment.

The application provides an image noise estimation method and system based on image edge characteristics. The image noise estimation method comprises: obtaining an original image, dividing the original image into a plurality of sub-images to be estimated; judging each sub-image to be estimated based on a gradient image of the sub-image to be estimated to determine whether the sub-image to be estimated is a valid sub-image to be estimated; performing convolution operation or correlation operation on the valid sub-image to be estimated by using a second-order difference filter kernel to obtain a second-order difference image; and estimating noise in the original image based on the second-order difference image to obtain a noise estimation value of the original image. The method obtains a sub-image to be estimated containing more noise and uses the valid sub-image to be estimated to estimate the noise of the original image, thereby accurately estimating the distribution of the noise in the original image. The system can realize the above method.

The invention discloses an offshore wind plant underwater noise real-time estimation method and device, a medium and a product, and relates to the field of underwater noise estimation.The method comprises the steps that an underwater noise sound pressure signal, a three-axis vibration acceleration time domain signal, an air sound pressure time domain signal, the wind speed, the cabin rotating speed and active power output are collected; segmenting each time domain signal, performing fast Fourier transform on each time domain signal in each time window, estimating power spectral density, dividing the power spectral density into a plurality of frequency bands, and calculating an underwater sound pressure level, a structural vibration acceleration level and an air sound pressure level of the frequency bands according to the power spectral density; training in each operation condition interval to obtain a final inversion coefficient set corresponding to each frequency band in each operation condition interval; and according to the operation condition interval to which the current environment condition belongs and the inversion coefficient set of the frequency band, calculating the underwater sound pressure level estimation value of each frequency band, and generating underwater noise spectrum distribution in a set frequency range, so that the underwater noise of the offshore wind plant can be efficiently and accurately estimated.

The invention discloses a method for training a de-noising model for de-noising an image of a scanning electron microscope and a related product. The method comprises the following steps: acquiring a scanning electron microscope image and label noise data; performing noise estimation on the electron microscope image to obtain a noise estimation result; dividing the scanning electron microscope image into different noise intervals according to the noise estimation result; defining a first label and a second label based on a noise difference between a noise estimation result corresponding to each scanning electron microscope image in the corresponding noise interval and the corresponding label noise data; and combining the scanning electron microscope images of the corresponding noise intervals with the corresponding first labels or second labels to train a plurality of denoising models, and obtaining target denoising models adapted to different noise intervals. By using the scheme of the invention, accurate balance between a noise reduction effect and detail reservation can be realized.

The invention belongs to the technical field of biomedical engineering, and discloses a gene regulatory network optimization method based on a diffusion model, which comprises the following steps: acquiring gene data of cells under a steady state condition, and constructing a gene expression matrix according to the gene data; injecting Gaussian noise into the gene expression matrix based on a diffusion model method to generate a series of noisy data sequences; performing noise estimation and structure estimation on the noisy data sequence by a noise estimator and a structure estimator based on a gene regulation and control network, and performing reverse denoising processing according to the noise estimation and the structure estimation to obtain gene structure estimation after reverse denoising; performing structure optimization on the gene structure estimation after reverse denoising by adopting an acyclic constraint function and a regularization substitution method; and outputting the optimized gene structure estimation. According to the method, the regulation and control relation between the genes is accurately recognized from high-dimensional gene expression data, and the modeling precision of the regulation and control relation between the genes is improved.

This application proposes a ground-penetrating radar (GPR) data augmentation method and apparatus. First, random Gaussian noise and specified category labels are obtained. Then, the random Gaussian noise and specified category labels are input into a trained conditional latent diffusion model. An accelerated sampling strategy is used to generate GPR data samples of the corresponding categories. The conditional latent diffusion model includes a variational autoencoder and a noise estimation network. The variational autoencoder includes an encoder and a decoder; the former maps the data to the latent space, and the latter reconstructs the data from the latent space. The noise estimation network performs condition-guided denoising processing in the latent space. This allows for the stable and efficient generation of GPR data samples for multiple target categories.

The invention provides a hyperspectral and laser radar fusion classification method (GGCDM) based on gating guide condition diffusion, which is used for fusing hyperspectral image (HSI) and laser radar (LiDAR) data to realize high-precision ground feature classification. The existing method is difficult to consider deep coupling of high-dimensional spectral features and three-dimensional geometric information under the problems of insufficient multi-modal feature interaction and limited generalization ability. According to the method, by designing a gating condition modulator (GCM), HSI and LiDAR features are mapped to a unified potential space, and an interactive perception gating structure is introduced to realize dynamic weight adjustment of two modal features, so that the cross-modal cooperative characterization capability is enhanced. Meanwhile, a deep interaction enhancement module (DIEM) is embedded in a diffusion reconstruction network, cross-modal association in a potential space is explicitly modeled, and the stability and robustness of a classification result are improved. Different from a traditional diffusion model based on noise estimation, the method adopts an image reconstruction normal form, takes a classification graph as a generation target, avoids training instability, and remarkably improves generalization performance. Experimental results show that the classification precision of the method is superior to that of an existing method on multiple groups of real data sets. The method can be widely applied to the fields of remote sensing image intelligent interpretation, land cover classification, environment monitoring and the like.

The application provides an iterative-based image noise estimation method and system. The method comprises: obtaining a gradient amplitude image of an original image and an initial noise estimation value of the original image, and obtaining an initial binarization threshold based on the initial noise estimation value; performing binarization processing on the gradient amplitude image based on the binarization threshold to obtain a binarization region; calculating a gradient amplitude average value of the gradient amplitude image corresponding to the binarization region; updating the binarization threshold based on the gradient amplitude average value; sequentially repeating the binarization processing step and the binarization threshold updating step until a preset stop iteration condition is met; and taking the latest binarization threshold as the noise estimation value of the original image. The method is to firstly set or obtain an initial noise estimation value of an image, perform binarization processing on the gradient amplitude image, update the binarization threshold, and use the binarization threshold updated continuously to approximate the most accurate estimation of the noise in the image.

The application discloses a method for estimating unknown state and noise of a nonlinear system, which comprises the following steps: constructing a general nonlinear system model with measurement noise and sampling output; constructing an extended nonlinear system through filtering transformation; and proposing an extended non-fragile observer to estimate unknown state and noise of the constructed extended nonlinear system. The proposed non-fragile observer is verified, and it is proved that the non-fragile observer can still track unknown state of the nonlinear system under the conditions of noise and gain disturbance. The proposed observer can still well track unknown state of the extended system.

A hearing device includes input transducers providing a transducer input, the input transducers comprising a first input transducer for provision of a first transducer input signal, wherein the transducer input is based on the first transducer input signal; a processor configured to provide an electrical output signal based on the transducer input; and a receiver to provide an audio output signal based on the electrical output signal; wherein the processor is configured to process the transducer input for provision of a first input and a second input; apply a neural network to the second input for provision of a second output that is a prediction of future noise; provide a noise estimate based on the second output; and subtract the noise estimate from a magnitude of the first input for provision of a first output, wherein the electrical output signal is based on the first output.

The invention discloses an unmanned aerial vehicle multi-source fusion navigation positioning method independent of satellite signals, and belongs to the technical field of unmanned aerial vehicle docking navigation, and the method comprises the following steps: 1, building a fusion navigation positioning system; 2, performing time registration and outlier elimination; 3, solving an information distribution coefficient; 4, time-variable measurement noise estimation is carried out; 5, updating time and measurement; step 6, fusion and differential feedback of the main filter; according to the scheme, the problems of insufficient sensor redundancy, rigid multi-source fusion navigation algorithm information distribution and fault isolation mechanism and the like of the existing docking navigation positioning system are solved, so that the navigation positioning precision and the anti-interference capability of the unmanned aerial vehicle are improved.

Artificial intelligence for time-series data analytics is provided. A first time-series data set is provided to a pre-trained recurrent neural network trained based on a second time-series data set. A prediction of a ground truth state of the first time-series data set is received therefrom. The first time-series data set is provided to a dynamical recurrent neural network trained based on the second time-series data set and the pre-trained recurrent neural network. A noise-reduced prediction of a ground truth system state of the first time-series data set is received therefrom. An estimate of sensor noise is read. The estimate of sensor noise is generated based on the second time-series data set and the pre-trained recurrent neural network. A prediction of a state of the system is generated based on the pre-trained recurrent neural network, the dynamical recurrent neural network, and the estimate of sensor noise.

The invention provides a real-time speech enhancement method and system based on dual-stage spectral subtraction and dual-mask fusion, and relates to the technical field of speech signal processing, and the method comprises the steps: respectively carrying out the framing, windowing and short-time Fourier transform of a left channel mixed signal and a right channel noise reference signal, obtaining a complex frequency spectrum and an amplitude spectrum of the left channel signal and an amplitude spectrum of the right channel noise reference signal; and performing noise estimation by adopting a first noise multiplication factor based on the amplitude spectrum of the right channel noise reference signal to obtain a noise estimation spectrum, and performing constraint spectrum subtraction on the amplitude spectrum of the left channel signal to obtain voice amplitude estimation of a first stage. According to the method, effective suppression of TTS noise and real-time speech enhancement are realized through framing windowing and frequency domain conversion in combination with over-estimation spectrum subtraction and double-mask fusion through two-stage gain application and time domain reconstruction.

The invention provides a signal processing method and device, electronic equipment, a chip and a storage medium, and relates to the field of communication, and the method comprises the steps: carrying out the noise estimation of a receiving signal received by at least one receiving antenna, so as to obtain an initial noise correlation matrix of the receiving signal; according to the initial noise correlation matrix, determining an interference scale factor used for representing the proportion of white noise and interference in the received signal; according to the interference scale factor, correcting the initial noise correlation matrix to obtain a target noise correlation matrix; and performing noise and interference suppression on the received signal based on the target noise correlation matrix to obtain a demodulation signal. Therefore, the difference between the estimated noise correlation matrix and the real noise correlation matrix can be reduced, the accuracy of noise correlation matrix estimation is improved, and on the basis, noise and interference suppression is performed on the received signal based on the accurate noise correlation matrix, so that the finally obtained demodulation signal is closer to the real sending signal.

This paper introduces procedures for estimating noise covariance across resource blocks with similar noise distributions. These procedures lead to more accurate estimates of noise occurring in a given channel because accuracy can be improved by increasing the number of resource blocks examined, while also identifying and filtering out those contaminated by interference. In summary, these procedures represent an automated method for detecting imbalances between noisy resource blocks and resource blocks that also contain interference, then forming clusters of resource blocks with similar characteristics to provide more samples that can be used to estimate the noise covariance.