29results about How to "Solve the noise problem" patented technology

The invention relates to an underwater image detection system and method based on a YOLOv8s model, and the system comprises an underwater image collection and preprocessing subsystem which is used for collecting underwater image data and carrying out the preprocessing; the target detection subsystem is used for carrying out target detection by adopting an improved YOLOv8s model, and the improved YOLOv8s model is used for obtaining a feature map by adopting a mode of combining a denoising mechanism and a dynamic adjustment receptive field mechanism based on the preprocessed underwater image data, carrying out feature enhancement by adopting a dual attention mechanism based on the feature map, and carrying out target detection by adopting an improved YOLOv8s model; performing feature fusion and detection to obtain a target detection result; the performance monitoring and evaluation subsystem is used for monitoring and evaluating the target detection result; and the detection result display and output subsystem is used for outputting and displaying the target detection result. Compared with the prior art, the method has the advantages of improving the depth and precision of feature extraction, improving the underwater image detection performance and the like.

The invention provides a multi-view clustering method based on uncertainty and a graph neural network, and relates to the technical field of multi-view clustering. The method comprises the following steps: obtaining a multi-view data set, and preprocessing to obtain an original feature matrix and an adjacent matrix; performing feature extraction and fusion on the input original feature matrix and the adjacent matrix to obtain a consistent embedded matrix and a transfer matrix; based on a Dempster-Shafer evidence theory and Dirichlet distribution, obtaining reliable consistent embedding by reducing the uncertainty of an embedding space in the consistent embedding matrix; mapping the reliable consistent embedding and the transfer matrix into final embedding; and performing decoding reconstruction and clustering analysis on the final embedding to obtain a clustering result. According to the method, potential information in the views can be fully mined, the uncertainty in the model is reduced, the noise problem existing in a multi-view clustering task is solved, and the clustering result of the multi-view data can be quickly, efficiently and accurately obtained.

The invention discloses a rare earth mining area land utilization fine classification method for hyperspectral remote sensing images, and belongs to the field of remote sensing. Based on a hyperspectral remote sensing image of a rare earth mining area, an object-oriented feature extraction module and an object-oriented one-dimensional convolutional neural network classification model based on a multilayer attention mechanism are designed, and a rare earth mining area land utilization classification task is completed through the two modules. By combining the object-oriented and convolutional neural network, the classification precision of different land utilization categories is enhanced, and the phenomenon of salt and pepper noise in the classification result is avoided at the same time.

The invention discloses a Tibetan language Andole speech enhancement method suitable for specific noise and general noise, and relates to the technical field of low-resource language speech processing. The method specifically comprises the following steps: acquiring noisy voice information; preprocessing the noisy voice information, and dividing the processed data into a training set and a test set; a noise spectrum feature input by the noise classification module is optimized based on a weighted average method of a logarithmic power spectrum domain hard threshold; constructing a noise classification task and noise suppression task fused multi-target Tibetan speech enhancement model; training and testing the built speech enhancement model by using the training set and the test set to obtain a trained speech enhancement model; and inputting the actually collected noisy speech into the trained speech enhancement model to obtain an enhanced logarithmic power spectrum. Through effective fusion of the noise classification task and the suppression task, the quality and the intelligibility of the Tibetan language Andor dialect enhancement voice are improved, the human-computer interaction of the Tibetan language Andor dialect is more perfect, and the method has a positive effect on popularization of Tibetan language informatization.

The utility model discloses an ice maker which comprises a pre-cooling water path plate, an evaporator and a spraying water path plate, the evaporator comprises an evaporation pipe and a plurality of ice fingers which are fixedly connected with the evaporation pipe and extend downwards, and the evaporation pipe is clamped between the pre-cooling water path plate and the spraying water path plate; the pre-cooling water path plate is provided with a raw water inlet, a pre-cooling cavity communicated with the raw water inlet and a pre-cooling water outlet communicated with the pre-cooling cavity, and the spraying water path plate is provided with a pre-cooling water inlet communicated with the pre-cooling water outlet, a spraying cavity communicated with the pre-cooling water inlet and a spraying opening communicated with the spraying cavity. The spraying waterway plate is further provided with a through hole for the ice finger to penetrate through, and the spraying openings are arranged around the through hole. The interlayer design of the pre-cooling water path plate, the evaporator and the spraying water path plate can protect the evaporator on one hand, and on the other hand, the cooling capacity of the evaporator can be fully utilized to pre-cool raw water, so that the ice making efficiency is improved.

The utility model relates to the field of dimming drive circuits, and particularly discloses an LED analog-digital hybrid dimming drive circuit, which comprises an MCU controller and a dimming drive chip, the low-side output end of the dimming drive chip is electrically connected with a power switch element, the power switch element is electrically connected with the negative electrode of an LED lamp through an inductance element, and the power switch element is electrically connected with the negative electrode of the LED lamp through the inductance element. The PWM output end of the MCU controller outputs digital signals to the PWM input end of the dimming driving chip, a digital-to-analog conversion integrating circuit is arranged between the IO output end of the MCU controller and the ADJ input end of the dimming driving chip, and the IO output end of the MCU controller outputs digital signals to the digital-to-analog conversion integrating circuit. The digital-to-analog conversion integrating circuit converts the digital signal into an analog signal and transmits the analog signal to the ADJ input end of the dimming driving chip.

The application discloses a false information detection method, device and equipment based on a propagation forest, and the method comprises the following steps: acquiring propagation data of each source blog in a blog propagation in a social network for a source blog set to be detected, wherein the propagation data comprises propagation content and propagation relationship; extracting text features of the propagation content; calculating a propagation prototype and constructing a propagation tree after dividing the source blog set into K class clusters according to the text features; constructing a propagation forest based on the propagation prototype and the propagation tree; aggregating node neighborhood representations in the propagation forest to obtain embedding representations of the nodes; calculating enhanced embedding representations of source blog nodes and subsequent propagation blog nodes according to corresponding text features and embedding representations respectively; and classifying the enhanced embedding representations of the source blog nodes and the subsequent propagation blog nodes by applying a detection model to obtain false detection results of each source blog in the source blog set. The application can improve the performance of a false information detection task.

This invention discloses a smart contract similarity detection method that integrates global and local features. It extracts bytecode function blocks using a contract function block extraction tool, encodes these function blocks into feature vectors using the SimHash algorithm, and calculates the similarity matrix of the function blocks using the Hamming distance algorithm, thus obtaining local similarity values. Next, it extracts the contract's program dependency graph by adding control flow and data flow analysis, and then encodes the program dependency graph into feature vectors using a graph embedding model encoder based on the Graph2vec algorithm. These feature vectors are used as input to a similarity calculation function to obtain the global similarity value of the contract. Finally, the similarity result of the contract is calculated by fusing global and local features. This invention's method of integrating global and local features considers multi-dimensional contract code features, achieving more accurate detection results and possessing good versatility and practical value.

The utility model discloses an integrated cooker with a noise reduction fan, which is characterized in that a volute is limited by arranging a first connecting piece, rigid collision between the volute and the first connecting piece is avoided by movement of the first connecting piece, and the problem that noise is generated by vibration of the volute in the prior art is solved. The first connecting piece adopts an arc baffle, surrounding plates are arranged on the two sides of the arc baffle, a limiting groove is formed between the surrounding plates, and the volute is limited in the limiting groove, so that the volute and the first connecting piece are mutually limited. The arc baffle is attached to the outer surface of the volute, so that rigid collision between the volute and the baffle is avoided again. The device is simple and easy to realize, and has good practicability.

The invention provides sound insulation cotton, a compressor and an air conditioner, relates to the technical field of compressors, and solves the technical problems that a sound insulation cotton structure is poor in low-frequency noise suppression effect, narrow in noise reduction bandwidth, complex in process and high in cost. The sound insulation cotton comprises a low-frequency noise control layer and a medium-high frequency noise control layer. The low-frequency noise control layer comprises a substrate and a resonance unit; the base is made of a cork material; the resonance units are arranged on the substrate in a periodic arrangement mode; and the medium-high frequency noise control layer is made of an environment-friendly porous fiber sound insulation cotton material. The sound insulation cotton contains the periodic metamaterial units, negative equivalent mass density is formed under the specific frequency, a low-frequency elastic wave band gap is achieved, the obvious inhibition effect on low-frequency noise below 500 Hz is achieved, and the broadband noise reduction effect can be achieved by adhering a common sound insulation cotton material; in addition, the device has the advantages of being simple in structure, small in overall thickness, easy to machine and install, wide in material source, low in cost, environmentally friendly and wide in applicability.

This utility model discloses a low-noise water heater, comprising: a heating element disposed at the bottom of a cavity; the heating element includes a heating tube and a heat spreader plate, the heating tube and the heat spreader plate being thermally connected, the heat spreader plate being attached to the bottom of the cavity, and the heating tube being spirally arranged; the ratio of the thickness of the heat spreader plate to the surface load of the heating tube is greater than or equal to 40%, and the thickness of the heat spreader plate is 3.5-5mm; wherein, the effective heat transfer area ratio of the heating tube and the heat spreader plate is greater than or equal to 50%, and the heating tube and the heat spreader plate cooperate to homogenize the heat generated by the heating tube; a first through hole is provided on the lower surface of the cavity, and a water inlet pipe passes through the spiral gap of the heating tube and the first through hole in sequence; a bracket is provided on the inner surface of the cavity, the bracket being disposed on the first through hole, and the bracket being used to buffer the water flow of the water inlet pipe. This device fundamentally solves the noise problem caused by uneven heating and bubble bursting during the heating process.

The invention provides a water pan structure and an air conditioner with the water pan structure, the water pan structure comprises a barrel, the barrel is provided with an inlet end and an outlet end which are oppositely arranged in the axial direction of the barrel, the outlet end of the barrel is provided with a protruding part, and the protruding part extends in the direction away from the inlet end in the axial direction of the barrel; partition pieces are arranged in an inner cavity of the barrel body and extend in the axial direction of the barrel body so as to partition the inner cavity of the barrel body. The problem that in the prior art, noise is generated in the refrigeration process of an air conditioner is solved.

The present invention provides a multi-view clustering method based on uncertainty and graph neural network, which relates to the technical field of multi-view clustering. The present invention obtains a multi-view data set, performs preprocessing, and obtains an original feature matrix and an adjacency matrix; extracts and fuses the input original feature matrix and the adjacency matrix to obtain a consistent embedding matrix and a transfer matrix; based on Dempster‑Shafer evidence theory and Dirichlet distribution, a reliable consistent embedding is obtained by reducing the uncertainty of the embedding space in the consistent embedding matrix; the reliable consistent embedding and the transfer matrix are mapped to a final embedding; the final embedding is decoded, reconstructed and clustered to obtain a clustering result. The present invention can fully tap the potential information in the view, reduce the uncertainty in the model, solve the noise problem existing in the multi-view clustering task, and can quickly, efficiently and accurately obtain the clustering results of multi-view data.

The invention discloses a system anomaly detection method for adaptively fusing multi-dimensional sequence data. The method comprises the following steps: firstly, collecting original monitoring data related to time in a system execution stage, carrying out digital coding on the original monitoring data in a non-digital format, then carrying out normalization processing to obtain a multi-dimensional time sequence # imgabs0 #, then reconstructing the multi-dimensional time sequence # imgabs1 # through a time sequence reconstruction model, and calculating a data difference value before and after reconstruction, and a reconstruction difference value sequence X is obtained. And constructing a feature space for the data distribution of the reconstructed difference sequence X, and repeating for t times. And calculating a feature vector of the reconstructed difference sequence X based on the feature space constructed by t times of experiments. Dividing the feature vector of the reconstructed difference sequence X into a plurality of subsequences by using a sliding window with a fixed length; and finally, calculating the similarity between different subsequences by using kernel mean value embedding, and judging the time corresponding to the subsequences of which the similarity is smaller than a threshold value as the abnormal operation moment of the system. And multi-dimensional time series data are effectively utilized.

The invention provides a contrast enhancement fusion electrocardio large model training deployment method and system. The method comprises the steps that S1, electrocardio data and inquiry information are collected; s2, the electrocardio data are preprocessed, and preprocessed electrocardio data are obtained; s3, extracting a structured text tag from the preprocessed electrocardiogram data, and constructing an enhanced electrocardiogram data set; s4, using an enhanced electrocardiogram data set to compare, learn and train the model; s5, analyzing the preprocessed electrocardiogram data and inquiry information by using the trained model, and outputting an inquiry result; and S6, deploying the multi-modal electrocardio basic model to a cloud server with a GPU (Graphics Processing Unit). According to the method, the signal-language-label comparative learning normal form is adopted, so that the problem that the representation learning accuracy is affected due to the fact that a noise positive example is introduced into a multi-label classification task in traditional comparative learning is solved, the representation of the model in a complex classification task is enhanced, and the method has a good application prospect.

The present application discloses a line of sight detection method, device, equipment and medium. The present application obtains a facial image of a target object, determines the head posture of the target object according to the facial image of the target object, and if the yaw angle and the pitch angle in the head posture meet preset conditions, performs three-dimensional line of sight analysis processing on the facial image to determine the line of sight of the target object. When the yaw angle and the pitch angle are within a preset angle range, the accuracy of line of sight detection can be guaranteed and the efficiency of line of sight detection can be improved. If any one of the yaw angle and the pitch angle does not meet the preset conditions, performs line of sight analysis processing on the facial image using a preset line of sight recognition model to determine the line of sight of the target object. This can solve the noise problem caused by sample imbalance when training the preset line of sight recognition model and improve the accuracy of line of sight detection.

The invention provides a federal construction method of a medical data classification model, which is used for combining a plurality of labeled clients and a plurality of unlabeled clients to construct the medical data classification model, and is characterized in that the labeled clients indicate that a data set on the clients comprises a plurality of medical data and corresponding disease labels; a label-free client indicates that a data set on the client only comprises a plurality of medical data. The method comprises the following steps: a federated client clustering step is used for clustering clients with similar data distribution into a sub-federated set; the federal semi-supervised self-training step is used for sorting the clients in each sub-federal set, and labeling the non-label clients according to a sorting result so as to enable all medical data of each non-label client to be endowed with target false labels; and the federated client training step is used for executing multiple rounds of federated training based on all the labeled clients and all the label-free clients endowed with the target pseudo labels to obtain a medical data classification model.

The invention provides a 3D scene arbitrary angle model segmentation method based on 3D Gaussian Splatting, which belongs to the field of machine perception technology and includes point cloud data extraction, 3D model training, model category feature extraction, model category feature training, model segmentation, and noise elimination; the system uses Colmap to reconstruct dense 3D point cloud information and camera trajectory from an image sequence; uses a 3DGS method to perform 3D reconstruction on the point cloud information; uses a DEVA model to perform image segmentation and image category preprocessing; trains a multilayer perceptron composed of multiple fully connected layers as a discriminator for segmenting the 3D model, which greatly reduces the discriminator training overhead; uses a KNN algorithm to calculate the average distance between each Gaussian kernel and several adjacent Gaussian kernels to judge the noise, so as to solve the detail problem that occurs during model segmentation, and can also solve the object boundary roughness problem; finally, the 3D model arbitrary angle segmentation method based on 3DGS is realized.

A multi-purpose valve and its valve plate and application, wherein the valve plate includes an elastic sheet structure that is concave inward; the valve includes a cylinder composed of three valve plates of the same structure, the concave part of each valve plate is in close contact with the concave parts of the other two valve plates, thereby achieving sealing inside the cylinder, and the outer edge of the valve plate is embedded in the bracket to form a cylinder supported by the bracket, which is installed in the inner tube, and one end is plugged with an inlaid cover to form a negative pressure valve core; the inner tube is installed with a sealing ring and pressed into a valve shell with a duct seat on the side; the valve and valve plate can be used in sewage valves. The present invention can directly replace the existing negative pressure valve, partially directly replace the pipe clamp valve, and can also replace the slide valve and butterfly valve for the newly built feces collection system. It is a truly multi-purpose valve. The present invention has a simple structure, small size, light weight, and reliable performance. It can greatly reduce the maintenance workload, reduce the application cost, and can be quickly replaced, which can effectively reduce the cost of the high-speed rail feces collection system.

This application relates to the field of intelligent transportation and proposes a calibration method, device, equipment, and storage medium for roadside radar. The method includes: acquiring first positioning information and speed information of a vehicle, and second positioning information of a connected vehicle; clustering the first positioning information to generate a vehicle trajectory; determining the vehicle trajectory matching the second positioning information based on the relative positional relationship of the vehicle trajectories and pre-determined lane information of the connected vehicle; aligning the second positioning information and third positioning information in the vehicle trajectory according to time, and performing coordinate transformation calculations on the aligned positioning information to determine the calibration information of the roadside radar. Because this application generates vehicle trajectories by clustering based on the first positioning information and speed information, it can effectively solve the noise and discontinuity problems of roadside radar, and it eliminates the need to close roads when calibrating roadside radar on public roads.

The invention discloses a stomach intestinal metaplasia grading method and system based on noise robustness and an open set, and belongs to the technical field of medical image processing. Processing the stomach endoscope image based on a pre-trained intestinal metaplasia grading model, and outputting an intestinal metaplasia grading result; the intestinal metaplasia grading model comprises a noise identification module, a comparison learning module, a label correction module and a classification output module; the pre-training process comprises the following steps: obtaining a clean sample set and a noise sample set based on a noise identification module; calculating class prototype similarity based on a comparative learning module, and discriminating valid noise and invalid noise; calculating an effective noise confidence coefficient based on a label correction module, and generating a differentiated pseudo label; and performing multi-module cooperative training based on the joint optimization total loss function, and outputting an intestinal metaplasia classification result through a classification output module. According to the method, the robustness of the model in noise and open set scenes is effectively improved, and the problem of 0-2-level fine-grained classification of intestinal metaplasia is solved.

A correlation-based network security for network devices is disclosed. Correlations between multiple network devices are mapped based on telemetry data from the multiple network devices to determine correlated devices. Behaviors of the correlated devices are monitored based on telemetry data received from the correlated devices to determine abnormal devices among the multiple devices. Priority alerts are generated for multiple network devices based on security alerts received for the abnormal devices.

This invention discloses an air conditioner for use in the home appliance industry, comprising: an energy storage device for containing energy storage material; a jet driving device assembled in the energy storage device; and a flow distribution device connected to the energy storage device via the jet driving device. When the jet driving device acts on the energy storage device, the energy storage device jets the energy storage material into the flow distribution device, and the jetted energy storage material is dispersed and emitted in the flow distribution device to release heat or cold energy. This invention solves the noise problem of air conditioners.

The invention relates to buried anti-flooding digital water supply pressurizing equipment which comprises a closed barrel, at least one shaft is arranged in the closed barrel, and the shaft comprises a water inlet transverse pipe, a vertical pipe and a water outlet transverse pipe which are connected in sequence. A submersible pump used for pressurizing fluid from the water inlet transverse pipe and outputting the fluid to the water outlet transverse pipe is arranged in the vertical pipe, a sealed container is arranged in the water inlet transverse pipe, and a control unit used for controlling starting and stopping of the submersible pump is arranged in the sealed container. By arranging the closed barrel, the booster pump station can be buried underground, and the occupied space of the ground is saved; the submersible pump is used as a power source, and the noise problem is solved; the control unit is shrunk and hidden in the water outlet parallel connection pipeline, the control unit can be cooled through flowing of water, heat is rapidly taken away, and it is guaranteed that the control unit operates in a low-temperature state all the time.

The utility model relates to pipeline sound attenuation technical field, concretely relates to a kind of rhombic small hole pipeline silencer, including outer sleeve and sound attenuation component, sound attenuation component includes top sealing plate, bottom sealing plate, air outlet, air inlet, sound attenuation cylinder and sound attenuation cylinder cover plate, air inlet and air outlet are installed with pipeline, only need to reserve the installation length and space of overall silencer on site, sound attenuation cylinder is steel structure type, without sound attenuation cotton setting, without replacing sound attenuation cotton, reduce vibration, noise through the structure design of prismatic sound attenuation hole, improve sound attenuation effect, sound attenuation cylinder cooperates multiple prismatic sound attenuation hole, and the sound attenuation effect of circular small hole silencer is good, and compared with circular small hole, noise reduction is improved by 3-5dB, to generate local eddy current at rhombic corner, and the attenuation of pipeline low-frequency vibration and airflow pulse vibration plays an active role, to solve the noise problem generated by low-frequency pipeline airflow vibration.

A display device is disclosed, comprising: a display panel including pixels defined by intersections between gate lines and data lines formed in a pixel area; a data driver, which sequentially outputs data signals in each sub-horizontal period in a horizontal period and outputs a selection signal including a transistor conduction period corresponding to the sub-horizontal period; and a distributor, which includes a transistor connected to each of the data lines and is switched according to the selection signal to output the data signal sequentially output from each source channel to the connected data line, wherein the transistor conduction period in the selection signal is different from the sub-horizontal period.

The invention discloses a compressor pipeline self-adaptive expansion silencer and a control method thereof.The compressor pipeline self-adaptive expansion silencer comprises a compressor body, the compressor body is connected with a self-adaptive expansion cavity through a compressor gas conveying pipe, the self-adaptive expansion cavity can change the length of the cavity and the sectional area of the cavity, and a sensor integration module is arranged in the compressor gas conveying pipe; the sensor integration module is connected with a central processing unit and a water circulation hub, the water circulation hub is connected with the self-adaptive expansion cavity, and the water circulation hub, the sensor integration module and the self-adaptive expansion cavity are all in communication connection with the central processing unit. The invention has the characteristics of automatically adjusting the silencing frequency of the expansion cavity and automatically adjusting the transmission loss of the expansion cavity.

The application relates to the technical field of intelligent application of automobiles, in particular to an automatic driving perception method and device of a vehicle, a vehicle and a storage medium, wherein the method comprises the following steps: acquiring initial picture data of the surrounding environment of the vehicle; inferring at least one target task according to the initial picture data, and detecting whether preset adversarial samples exist in the initial picture data in the inference process; if the initial picture data does not have the preset adversarial samples, obtaining a perception result of the initial picture data after the inference is completed; otherwise, the preset adversarial samples of the initial picture data are denoised by using a preset denoising strategy, first picture data after denoising is obtained, and the perception result of the initial picture data is obtained again based on the first picture data. Therefore, the problems that all data are usually denoised in the related art, the workload of data processing is large, the perception efficiency is low, and noise of input data of a perception task cannot be effectively avoided are solved.

The present invention provides a disconnect device, in which the disconnect device comprises: a housing having a support ring and a one-side half axle gear therein, the one-side half axle gear meshing with a pinion gear inside the support ring; a first clutch ring connected to the sleeve and having a first clutch portion on a side facing the support ring; and a second clutch ring inserted into the first clutch ring, having a second clutch portion on one side facing the one side half axle gear, and coupled to the sleeve on the outer diameter portion on the other side.