11 results about "Convolution filter" patented technology

The application provides a point spread function fast spatial convolution method, which comprises the following steps: step 1, calculating an imaging profile and a point spread function field; step 2, determining the spatial position coordinates of the center position of the point spread function by circulation; step 3, determining an interpolation method and calculating the interpolation coefficients of the local imaging profile which contributes to the point spread function; step 4, performing wave number domain fast spatial convolution; step 5, multiplying the profile by the corresponding interpolation coefficients after filtering; and step 6, accumulating the filtering results of the local imaging profile to obtain a final filtered profile. The point spread function fast spatial convolution method can greatly improve the calculation efficiency of the point spread function spatial convolution filtering process and reduce the calculation cost based on the linear accumulation characteristics of the interpolation method.

The technology disclosed relates to constructing a convolutional neural network-based classifier for variant classification. In particular, it relates to training a convolutional neural network-based classifier on training data using a backpropagation-based gradient update technique that progressively match outputs of the convolutional network network-based classifier with corresponding ground truth labels. The convolutional neural network-based classifier comprises groups of residual blocks, each group of residual blocks is parameterized by a number of convolution filters in the residual blocks, a convolution window size of the residual blocks, and an atrous convolution rate of the residual blocks, the size of convolution window varies between groups of residual blocks, the atrous convolution rate varies between groups of residual blocks. The training data includes benign training examples and pathogenic training examples of translated sequence pairs generated from benign variants and pathogenic variants.

The embodiment of the application discloses a shadow generation method and related device, the method comprises the following steps: obtaining the depth data, normal data and illumination data of each point in a virtual scene as the shadow reference data of each point; inputting the shadow reference data of each point into a first convolution filter to obtain the basic shadow data of each point output by the first convolution filter; determining the half-shadow intensity of each point according to the distance from each point to a virtual light source in the virtual scene and the propagation distance of a virtual light emitted by the virtual light source; weighting the basic shadow data of each point by using the half-shadow intensity of each point to obtain the weighted shadow data of each point; and the weighted shadow data is used for shadow rendering for the points. The method can improve the setting efficiency of the shadow effect, and reduces the required artificial cost and time cost.

Systems and methods for image processing are described. Embodiments of the present disclosure obtain a low-resolution image and a text description of the low-resolution image. A mapping network generates a style vector representing the text description of the low- resolution image. An adaptive convolution component generates an adaptive convolution filter based on the style vector. An image generation network generates a high-resolution image corresponding to the low-resolution image based on the adaptive convolution filter. 20 23 28 57 07 18 D ec 2 02 3 A B S T R A C T 2 0 2 3 2 8 5 7 0 7 1 8 D e c 2 0 2 3 1 / 17 FIG. 1 120120 110110 105105 115115 125125 “a cute corgi lives in a house made out of sushi” low-resolution image high-resolution image 100100 1 / 17 115 125 high-resolution image 120 "a cute corgi lives in a house made out of +sushi" low-resolution image 110 105 100 FIG. 1 20 23 28 57 07 18 D ec 2 02 3 1 8 D e c 2 0 2 3 1 1 5 2 0 2 3 2 8 5 7 0 7 1 2 0 + 1 1 1 1 0 1 0 0

A method for deblurring distortion-corrected images, includes: obtaining a video-see-through (VST) image of a real-world environment; determining a given region of the VST image, based on at least one of: a distortion profile of a camera lens of at least one VST camera, blur characteristics of the camera lens, a scaling ratio between a resolution of the at least one VST camera and a resolution of a display whereat the VST image is to be displayed; and performing an undistortion deblurring operation on the given region of the VST image, by utilising at least one of: (i) a deblurring deconvolution filter, (ii) at least one neural network, based on locations of pixels of the given region, to generate an output image.

The application discloses a kind of self-adapting micro-doppler corner point feature extraction method and device, comprising: it is proposed to use micro-doppler corner point feature to realize through-wall radar human behavior recognition, and gives a kind of self-adapting corner point feature extraction method based on Gauss difference convolution filter and deformable convolution network;Micro-doppler corner point feature is defined as the point that gray scale sharply changes in different directions in radar square distance-time image and square doppler-time image, reflects the inflection point of human body limb node's motion trajectory curve, stationary point, curve intersection and boundary;The corner point feature extraction method proposed uses Gauss difference convolution filter to extract micro-doppler corner point supervision label on simulation data, then trains μD-CornerDet model using these labels;Inference stage, only μD-CornerDet model is used to obtain corner point feature map for measured data;The application verifies the effectiveness and robustness of the proposed method through numerical simulation and measurement.

A method for segmenting an image by using each of a plurality of weighted convolution filters for each of grid cells to be used for converting modes according to classes of areas is provided to satisfy level 4 of an autonomous vehicle. The method includes steps of: a learning device (a) instructing (i) an encoding layer to generate an encoded feature map and (ii) a decoding layer to generate a decoded feature map; (b) if a specific decoded feature map is divided into the grid cells, instructing a weight convolution layer to set weighted convolution filters therein to correspond to the grid cells, and to apply a weight convolution operation to the specific decoded feature map; and (c) backpropagating a loss. The method is applicable to CCTV for surveillance as the neural network may have respective optimum parameters to be applied to respective regions with respective distances.

This invention provides a continuous decoding method for electroencephalogram (EEG) signals based on a time-frequency dual-stream neural network, comprising: acquiring a window of the EEG signal to be decoded; inputting a dual-stream network consisting of a frequency domain branch, a time domain branch, and a fusion classification module; the frequency domain branch performing a Fast Fourier Transform on the window, concatenating the real and imaginary parts of the complex spectrum, and generating frequency domain embedding features through channel expansion and frequency multi-layer convolution after channel expansion and recalibration via hybrid attention; the time domain branch employing multi-scale convolution filtering in parallel, and generating time domain embedding features through spatial compression and residual time-domain enhancement after spatial compression and residual time-domain enhancement; concatenating the two embedding features, outputting the instruction category probability by a multilayer perceptron, and taking the maximum value as the continuous decoding result. This invention extracts complementary time-frequency features in parallel, retains the phase and amplitude information of the complete complex spectrum, and enhances the discrimination robustness by combining a hybrid attention mechanism, effectively solving the problem of transient noise interference under short time windows, and achieving high-precision continuous decoding of high-frequency EEG signals.

The present invention relates to a method for determining the concentration of a chemical compound in a fluid, by means of an optical measurement system for measuring a light intensity spectrum, such that: (i) by means of a training set comprising a plurality of measured training light intensity spectra for a plurality of training fluids of predetermined concentration, a model is built to determine the concentration of at least one chemical compound in a fluid by training a one-dimensional convolutional neural network on the training set, the network being such that the number of convolutional filters per layer decreases with the position of the layer in the network; (ii) by means of the optical measurement system, at least one light intensity spectrum of the fluid is measured; (iii) the constructed model is applied to the measured light intensity spectrum, and the concentration of the chemical compound in the fluid is determined.Figure 2 to be published.