43 results about "Max pooling" patented technology

The invention relates to a mechanical fault diagnosis method and system based on deep learning. The method comprises the following steps: converting a multi-source time domain signal into a time frequency image through continuous wavelet transform, extracting features by using a primary feature encoder, and extracting cross-source common features through adversarial training of a shared feature discriminator; therefore, a gated multi-scale encoder is guided to enhance common feature expression, and deep fusion of multi-source features is realized through a cross multi-head attention network. Global average pooling and maximum pooling are synchronously carried out on the fused features to give consideration to overall and local information, and a comprehensive feature vector is formed; and finally, by means of a double-branch diagnosis network, the training loss of the multi-class network is dynamically weighted according to the prior probability output by the binary network, so that multi-source information is effectively fused under the condition of data imbalance, and the accuracy and robustness of fault classification are remarkably improved.

The invention discloses a semantic and structure preserving-based point cloud adaptive downsampling method and device, and belongs to the field of computer point cloud analysis and feature learning. Firstly, point cloud features are extracted, a local neighborhood is constructed, and semantic features and space coordinates of all points are obtained; counting the number of times of selection of the feature channels in the neighborhood based on maximum pooling, and obtaining a local importance score through normalization; through cross-neighborhood aggregation and in combination with spatial distance attenuation weight, geometric consistency is enhanced, and a global importance score is generated; a lightweight multi-layer perceptron is used for fusing semantic and spatial features to predict a comprehensive importance score, and key points are selected according to the score to form a down-sampling subset; and finally, a teacher-student self-supervised training framework is adopted, and a high-confidence-coefficient pseudo tag is generated through a teacher branch to guide student branch parameter optimization, so that efficient reasoning is realized. According to the method, the semantic information and the geometric structure of the point cloud can be effectively kept while the data volume is remarkably reduced.

The application discloses a ground three-dimensional laser scanning point cloud collection system and relates to the field of three-dimensional point cloud data processing of the ground; the size of data input is 10 square meters; a KPConv algorithm is improved; the characteristics of each input point are expanded; Max Pooling is used to aggregate the relative position and the Euclidean distance of the field points and the center point; Average Pooling is used to aggregate the field characteristics of each point extracted through the multi-layer KPConv into global characteristics; and a set of point cloud visualization websites is deployed by using WebGL.

The invention relates to the technical field of image processing, in particular to a 3D point cloud classification segmentation method based on dynamic edge convolution and residual double attention, which comprises the following steps: acquiring an image to be processed; a 3D-AGCN network is constructed; feature alignment is carried out by using a Transform encoder, feature groups are obtained through a plurality of EADEC, RDA and AMFE modules, the feature groups are connected in series, and then convolution compression is carried out to obtain shared features; the classification branch performs global maximum pooling and average pooling on the shared features, and outputs classification scores; and the segmentation branches repeat the shared features and the one-hot category vectors to each point and then splice the shared features and the one-hot category vectors, and a segmentation score is output. The method solves the problem that an existing method still needs to be improved in the aspects of effectively fusing local details and global contexts and optimizing model performance.

This invention discloses a UAV detection method based on multi-view feature fusion using residual networks, comprising: constructing multi-view data: obtaining the time-domain plot, short-time Fourier transform plot, continuous wavelet transform plot, and Wegener-Will distribution plot of the measured signal; constructing a ResNet34, including an input structure, an intermediate structure, and an output structure; the input structure processes the input data through convolution and max pooling operations; the intermediate structure consists of four similar structural layers, each consisting of multiple residual blocks, each residual block containing three convolutional layers and a shortcut connection; starting from the second structural layer, the initial residual block of each structural layer also contains an up-dimensional sampling structure; constructing a multi-view feature fusion network model based on residual networks, and outputting the UAV detection accuracy after multi-view feature fusion at the output end. This invention improves the UAV detection efficiency by fusing features from the multi-view data of the signal.

This invention discloses a method for reconstructing building point clouds based on a DGCNN model with an improved KNN algorithm. The method includes: normalizing the original building point cloud to be reconstructed to obtain normalized point cloud data; constructing a DGCNN network based on the improved KNN algorithm and training the DGCNN network to obtain a trained DGCNN model. The DGCNN network based on the improved KNN algorithm includes a spatial transformation layer, four graph convolutional layers, a max pooling layer, a first multilayer perceptron, and a second multilayer perceptron connected sequentially; and inputting the normalized point cloud data into the trained DGCNN model to obtain the corresponding prediction results. This invention utilizes the local update mechanism of the KD tree to efficiently and dynamically adjust the adjacency graph during network training, avoiding the high computational cost of reconstructing the entire search tree.

The application discloses a kind of real-time high-resolution portrait matting methods based on deep neural network, including obtaining training dataset, and marking generation training groundtruth alpha matte;Training data set is data enhanced;Network model is trained in step phase;Using trained network carries out matting.Through embedding ConvLSTM module in network configuration, using Max Pooling Indices, high-definition detail optimization is carried out using PRM, semantic segmentation task is added, and the core technology of high-precision real-time portrait matting is created, simultaneously, data set and data enhancement method are innovated, and training is carried out in stages, from simple to complex, from rough to fine, the training effect of algorithm is strengthened, the innovation and application of the three aspects interact, mutually unified, the performance and practicality of algorithm are comprehensively improved, and powerful technical support is provided for high-precision real-time portrait matting application.

This invention provides a deep spiking neural network method for ECG classification based on attention and integer training pulse inference, comprising the following steps: acquiring raw ECG signal data and preprocessing the raw ECG signal data; performing three rounds of convolution and corresponding max pooling on the ECG data features; performing block-based local self-attention processing on the ECG data features to obtain feature associations in local regions of the ECG data features; performing global self-attention processing on the ECG data features to obtain feature associations across the entire sequence of the ECG data features; performing two rounds of convolution and corresponding max pooling on the ECG data features; classifying the integrated ECG data features using a classification head, and outputting the ECG data classification result. This invention can effectively mine long-range dependency information of ECG data and retain shallow feature information through residual fusion, avoiding the gradient decay problem in deep networks, thereby enabling deep spiking neural network learning.

Methods, systems, and / or apparatuses are described for detecting relevant motion of objects of interest (e.g., persons and vehicles) in surveillance videos. As described herein input data based on a plurality of captured images and / or video is received. The input data may then be pre-processed and used as an input into a convolution network that may, in some instances, have elements that perform both spatial-wise max pooling and temporal-wise max pooling. The convolution network may be used to generate a plurality of prediction results of relevant motion of the objects of interest.

The embodiment of the specification provides a maximum pooling processing method and device for protecting privacy data. The method comprises the following steps: based on a local slice of an input matrix, a local slice of a first comparison matrix is determined by multiple parties, which indicates a comparison result of horizontally adjacent elements in the input matrix; based on the local slice of the first comparison matrix, a local slice of an intermediate result matrix is determined by multiple parties, an element of the intermediate result matrix being a larger value of horizontally adjacent elements in the input matrix; based on the local slice of the intermediate result matrix, a local slice of a second comparison matrix is determined by multiple parties, which is used for indicating a comparison result of vertically adjacent elements in the intermediate result matrix; and based on the local slice of the second comparison matrix, a local slice of a pooling result matrix is determined by multiple parties, an element of the pooling result matrix being a larger value of vertically adjacent elements in the intermediate result matrix. The communication overhead can be reduced, and the overall calculation efficiency can be improved.

This invention discloses a bridge bending detection method based on a lightweight multi-scale sparse gating network, relating to the field of bridge structural health monitoring. The method includes: acquiring multimode fiber speckle images corresponding to different bending states of the bridge and preprocessing them to obtain standardized speckle images; obtaining an initial feature map based on a lightweight multi-scale sparse gating network through initial convolutional layers and max pooling layers; extracting and fusing bending-sensitive features through a multi-level multi-scale feature fusion module to obtain a multi-scale fused feature map; using a learnable sparse gating module for feature selection to obtain a sparse enhanced feature map; modeling global dependencies through a global context enhancement module to obtain a globally enhanced feature map; and inputting the globally enhanced feature map into a dual-task prediction module to output the corresponding result. This method achieves synchronous, lightweight, and high-precision detection of bridge bending degree and location, effectively decoupling the problem of multi-parameter cross-sensitivity.

The present application relates to the field of computer vision and artificial intelligence, and specifically to the processing of the pooling layer of the convolutional neural network, and proposes a method and device for large-core pooling. The method mainly includes the HBLK / WBLK block mode of mapping the pooling output to the input, including: performing the operation of the pooling core kh / kw length characteristic data on the H / W direction dimension, saving the temporary results of the direction in the internal cache SRAM, storing the temporary data in the maximum output mode, and when the internal cache size range is exceeded, the H / W direction block of cblk is performed; the W and H directions represent the width and height directions of the feature map; kw / kh are the sizes of the W / H direction pooling core respectively. The device includes a pooling top layer control device, an input / output control device, an HBLK unit control device, an HBLK unit control device, and a pooling operation device. The present application can realize the operation of the pooling core with unlimited size, the hardware device proposed can achieve the balance of performance, bandwidth, power consumption and area, and can effectively solve the technical problems such as small pooling core size and high redundancy of tensor characteristic data operation in the prior art.

This application relates to the field of urban pipeline network analysis technology, and particularly to a pipeline network connectivity assessment method, device, and related equipment based on DS-PIGNN. The method includes: acquiring pipeline network data, constructing a macroscopic topology map, and collecting the axial microscopic physical field sequence of each pipeline; extracting the features of the microscopic sequence of each pipeline using a one-dimensional residual convolutional neural network, and obtaining a microscopic health state vector through global max pooling; injecting this vector into the edge features of the macroscopic topology map to drive a graph neural network trained with a physical constraint loss function (including Kirchhoff flow conservation), which dynamically calculates the attention weights between nodes based on the hydraulic features of the nodes and the microscopic health state vector; and performing information aggregation based on this dynamic weight, simultaneously outputting the pipeline failure probability and node connectivity reliability. This application achieves dynamic and coupled analysis of microscopic physical damage details and macroscopic network cascade failures, and ensures the physical reliability of the assessment results while maintaining computational efficiency.

This invention discloses a method and system for extracting feature points from point clouds based on a self-attention mechanism. The method includes: acquiring point cloud slices of a point cloud model; inputting the point cloud slices into a neural network to obtain multi-channel feature neighborhoods; performing MLP calculations on the multi-channel feature neighborhoods and then applying a self-attention mechanism to the calculation results to obtain global features; and sequentially performing max pooling, MLP, and FNN calculations on the global features to obtain the probability that the center point of the point cloud slice is a feature point. This invention obtains multi-channel feature neighborhoods from point cloud slices. In addition to the spatial location information of the point cloud, the multi-channel feature neighborhoods also include Euclidean distance information and center point neighborhood information, thus obtaining more semantic information. Combined with self-attention mechanism calculations and post-processing, the dimensionality of the output is reduced, decreasing the computational load of subsequent feature mapping; this allows for convenient and efficient acquisition of feature points.

The invention discloses a coarse-to-fine image matching method based on an aggregation attention mechanism, and belongs to the technical field of computer vision. The method comprises the following steps: firstly, extracting multi-scale features of an input image through a lightweight heavy parameterized convolutional neural network; then, an aggregation attention module is used for efficiently converting coarse-grained features, and the module is used for aggregating tokens through deep convolution and maximum pooling and enhancing feature discrimination ability in combination with rotation position coding; then calculating a similarity matrix based on the converted features, and obtaining rough matching point pairs through double softmax operation; and finally, by taking rough matching as guidance, realizing sub-pixel-level accurate positioning on fine-grained features through a two-stage process of mutual nearest neighbor screening and space expectation calculation. According to the method, the calculation efficiency is remarkably improved while the high matching precision is guaranteed, and the method is suitable for scenes such as unmanned aerial vehicle visual positioning and navigation which have strict real-time requirements.

This invention provides a small target detection method based on PBAF-YOLO with progressive boundary awareness and multi-scale fusion. The method involves inputting a feature map, calculating neuron energy values ​​based on the spatial mean and unbiased sample variance of each channel's feature values, generating spatial attention weights based on these neuron energy values, and then weighting and enhancing the input feature map to obtain boundary enhancement features. These boundary enhancement features are grouped along the channel dimension, and average pooling and max pooling are performed on each group in the horizontal and vertical directions respectively. These are then fused to generate direction-sensitive spatial attention weights, resulting in multi-scale enhancement features. Multi-scale feature maps output from different stages of the backbone network are reused, and feature fusion is performed through single downsampling and single upsampling paths with skip connections. The high-resolution features are iteratively optimized to generate a multi-scale detection feature map. This multi-scale detection feature map is then input into a high-resolution prediction head adapted for small targets, outputting the small target detection result.

The application discloses a kind of high-efficiency semantic segmentation networks, suitable for medical image analysis, automatic driving etc. The network adopts encoder-decoder architecture, combines CNN with Transformer (converter model), balances global modeling and computational efficiency. Encoder extracts multi-scale features through lightweight convolution, and introduces spatial selection module: its gate convolution splits channel into gate signal and reserved information, and key spatial features are activated and strengthened by Sigmoid activation function; Grouped pooling module extracts details using multi-scale pooling, and restores channels after upsampling and splicing. The decoder fuses multi-scale features and restores resolution through upsampling, enhances efficient channel attention, fuses global max pooling and average pooling, generates channel weight using one-dimensional convolution, and optimizes feature dependence. The design improves small target segmentation accuracy through gate mechanism and multi-scale pooling, is lightweight and easy to expand, and has high precision and practicality.

The invention discloses a data obfuscation method and device in collaborative maximum pooling calculation, a medium and a product, and relates to the field of data obfuscation, the method comprises the following steps: generating a first participant, and generating a private matrix by a second participant; the two parties respectively carry out dimension reduction on the private matrix sum to obtain the private matrix sum; the two parties respectively send the private matrix and the private matrix to each other, the first participant carries out maximum pooling on the addition result of the private matrix sum and then multiplies the addition result with the private matrix A to obtain a first pooling result, and the first pooling result is sent to the calculation requester; and the second participant performs maximum pooling on the addition result of the private matrix sum, multiplies the addition result with the private matrix B to obtain a second pooling result, and sends the second pooling result to the calculation requester.

To properly carry out learning with a convolutional neural network (CNN) while avoiding loss or disappearance of features of a target object.SOLUTION: A machine learning apparatus is configured to extract features of target data using CNN, to perform learning on the target data, the CNN including: a first convolutional layer configured to differentiate features of the target data; a second convolutional layer configured to increase the amount of information from the differentiated target data; a max pooling layer configured to compress the target data with increased amount of information, other than the maximum value; an upsampling layer configured to increase the size of the compressed target data for upsampling; a third convolutional layer configured to convert the upsampled target data to target data of a predetermined number of channels; and a fourth convolutional layer configured to directly convert the target data prior to the first convolutional layer to the target data of the predetermined number of channels and add the target data after the third convolutional layer and the target data after the fourth convolutional layer.SELECTED DRAWING: Figure 4

This invention discloses an automatic identification method for point cloud components in indoor scenes. The method first selects key points from the input point cloud data using farthest point sampling, performs structural ordering on the key points using a space-filling curve, and maps local blocks to the feature space based on a lightweight Point Net. Then, the features embedded by the mapping labels are encoded and decoded using a Transformer architecture. During downsampling pooling, the results of max pooling and average pooling are weighted and fused using learnable weight parameters. Simultaneously, Swing relative position encoding attention is introduced to capture the relative position information between points. Finally, the decoded features are processed through geometric topology awareness to integrate geometric accuracy and topological structure loss, thus completing the automatic identification of point cloud components. This invention provides support for subsequent analysis and research on 3D point cloud semantic segmentation.

The present disclosure relates to performing core crossing in hardware. The method is for receiving a request to process a neural network on a hardware circuit, the neural network including a first convolutional neural network layer having a stride greater than 1, and in response generating instructions that cause the hardware circuit to produce, during processing of an input tensor, a layer output tensor equivalent to an output of the first convolutional neural network layer by performing operations including: processing the input tensor using a second convolutional neural network layer having a stride equal to 1 or otherwise equivalent to the stride of the first convolutional neural network layer to produce a first tensor; zeroing elements of the first tensor that would not have been generated if the second convolutional neural network layer had the stride of the first convolutional neural network layer to produce a second tensor; and performing max pooling on the second tensor to produce the layer output tensor.

This invention provides a surface defect detection method and system based on multi-light source collaboration. The method includes: acquiring a set of images of an object under illumination from light sources in different directions; randomly cropping and adding noise to the training set samples to obtain training data; inputting a three-channel RGB image to each branch to obtain multi-branch feature information; constructing a network and using max pooling to fuse the multi-branch feature information, aggregating a variable number of feature vectors into a feature map with a fixed number of channels, retaining the most significant features of each branch to obtain a detection model to be trained; using the training data as input to the detection model to be trained, completing the model's forward computation to obtain a trained detection model; and applying the trained detection model to a real-world industrial quality inspection scenario to detect surface defects on workpieces. This invention solves the problem of difficult detection and classification of three-dimensional defects in industrial quality inspection scenarios by restoring the three-dimensional shape of an object from a two-dimensional image.

The application discloses a source code vulnerability detection method based on a graph neural network and multi-level attention fusion, and belongs to the technical field of intelligent software engineering and network space security. First, the source code is preprocessed and parsed to generate a code graph; a pre-trained model is used to extract graph structure features and graph node features; then, a graph neural network model is constructed, the graph node features are input into a multi-layer graph convolution network, and a self-attention mechanism is used to dynamically assign weights to each layer feature, and a global semantic feature matrix is generated through an average pooling layer; then, the original graph node features and the global semantic feature matrix are locally cross-fused, and a soft / hard attention fusion mechanism is used to extract local features; the fused features are input into a max pooling layer to generate local significant features; finally, a classifier is used to output a vulnerability detection result. The application uses global semantics to guide the extraction of local microscopic features, effectively improving the model's ability to capture and detect high-concealment code vulnerabilities.

A method and system for enhancing the performance of adversarial examples in retrieval question answering based on aggregated representation, the method comprising: S1: preprocessing the text of the question and answer to obtain the corresponding input text sequence X. q and X a S2: X q and X a Input a twin-tower model and obtain X. q and X a High-level representation vector H q and H a And syntactic representation vectors and S3: H q and H a The probability distribution mapped to the vocabulary is used to construct a representation vector v containing local lexical information through max pooling. q and v a S4: For v q and v a Pruning is performed to obtain a low-dimensional lexical representation vector. S5: The reduced-dimensional syntactic representation vector and the low-dimensional lexical representation vector are concatenated and aggregated to obtain the aggregated representation vector e. q and e a S6: According to e q and e a Calculate the matching score between question-answer pairs to obtain the matching loss function L. match S7: Based on L match Calculate the loss function L with added perturbation. adv The method of this invention can obtain more effective adversarial examples under adversarial training, thereby improving retrieval and question-answering performance.

This invention discloses a hyperspectral image classification method based on multimodal fusion. First, the dimensions of hyperspectral and lidar data are unified through a dimension unification layer. Then, multi-scale spatial features of the hyperspectral and lidar data are extracted through multi-scale grouped convolution, and multimodal feature aggregation is performed at the corresponding scales to obtain multi-scale aggregated features. These aggregated features are then concatenated to obtain multi-scale aggregated features, which are then subjected to global average pooling and global max pooling, and combined with global covariance information to obtain multi-scale global information of the multimodal data. Next, the multi-scale aggregated features at different scales are fused with the multi-scale global information of the multimodal data to obtain multimodal spatial fusion features. Finally, the spatial topology information of the lidar is used to impose topological constraints on the multimodal spatial fusion features, and the multimodal spatial fusion features under these constraints are input into a classifier to complete the classification task.

A method of training a machine learning artificial intelligence system that includes generating scenario realizations each having a virtual spatial layout of sound-influencing features, and generating acoustic recordings of sounds moving through each scenario realization, where each acoustic recording is based on propagation effects associated with a corresponding virtual spatial layout. The method may include identifying isolated sounds in the acoustic recordings, and training a machine learning model comprising a multi-layer convolutional recurrent neural network (CRNN), with the one or more isolated sounds, wherein the training is via rectified linear unit (ReLU) activation and max pooling along a frequency axis, wherein the trained machine learning model generates output event activity probabilities. The method may include receiving a subsequent acoustic recording of one or more subsequent sound sources, and classifying, via the trained machine learning model, the one or more subsequent sound sources based on the generated output event activity probabilities.

The invention relates to the technical field of computer application, and discloses a double attention-based supervised Chinese semantic matching method, which comprises the following steps of: firstly, acquiring an independent feature code and an associated feature code of a text pair by utilizing a double-channel coding strategy; then, parallel computing of affinity features and difference features is carried out by using a double attention module, and an affinity vector and a difference vector aggregating similar and conflict semantics are generated through an interactive alignment mechanism; performing self-attention enhancement and global maximum pooling on the vector by using a feature amplification module to obtain a global feature; meanwhile, performing linear mapping and auxiliary supervision training on the associated features by utilizing a fine tuning module, and extracting a fine tuning vector implying tag semantics; and finally, fusing the affinity global feature, the difference global feature and the fine tuning vector by using a multi-layer perceptron, and outputting a semantic matching result. According to the method, semantic consistency and conflict can be captured at the same time, feature discrimination is enhanced through label supervision, and the accuracy of text matching in a complex scene is effectively improved.

A method for training hotword detection includes receiving a training input audio sequence including a sequence of input frames that define a hotword that initiates a wake-up process on a device. The method also includes feeding the training input audio sequence into an encoder and a decoder of a memorized neural network. Each of the encoder and the decoder of the memorized neural network include sequentially-stacked single value decomposition filter (SVDF) layers. The method further includes generating a logit at each of the encoder and the decoder based on the training input audio sequence. For each of the encoder and the decoder, the method includes smoothing each respective logit generated from the training input audio sequence, determining a max pooling loss from a probability distribution based on each respective logit, and optimizing the encoder and the decoder based on all max pooling losses associated with the training input audio sequence.

The embodiment of the invention provides a power distribution area monitoring method and device, equipment, a storage medium and a program product. The method comprises the steps that a two-dimensional spectrogram can be generated by obtaining real-time monitoring data of a power distribution area, then the two-dimensional spectrogram is processed through a fault detection model, and a fault detection result can be obtained. The fault detection model comprises a maximum pooling layer, pooling can be carried out on a two-dimensional spectrogram, time domain and frequency domain data features are mined at the same time, weak signals in the early stage of a fault are effectively captured, and the problem of monitoring and early warning lag is solved. The fault detection model further comprises a hierarchical classifier, and the fault type can be accurately recognized according to the fault incidence relation, so that the fault recognition accuracy and response speed of power distribution area monitoring are improved.

A convolutional neural network may be embedded onto an integrated circuit (IC) device, which includes an embedder unit, a flow control unit, and etched mind unit(s). The embedder unit may generate a feature map from an input image. The etched mind unit(s) may be a hardware implementation of the CNN and execute neural network operations of the CNN using the feature map. An etched mind unit may include a convolution unit implementing convolution, a batch-norm unit implementing batch normalization, an activator unit implementing an activation function operation, a max pooling unit implementing max pooling, and an average pooling unit implementing average pooling, and a MatMul unit implementing matrix multiplication, each of which may has its own memory that stores weights or other data for performing a neural network operation. The flow contour unit may orchestrate the other components of the IC device based on a timing sequence of the network.