Abstract: An image processing method includes: obtaining a current image frame and a reference image frame; sequentially performing downsampling and upsampling on the current image frame to obtain a processed current image frame, and sequentially performing downsampling and upsampling on the reference image frame to obtain a processed reference image frame; according to a preset division manner, dividing the processed current image frame into current image sub-blocks and dividing the processed reference image frame into reference image sub-blocks; determining a reference image sub-block with a minimum similarity to each current image sub-block among the reference image sub-blocks as a matching block of the current image sub-block; obtaining a motion vector corresponding to the current image sub-block based on each current image sub-block and the matching block corresponding to the current image sub-block; and encoding the current image frame based on the motion vector.

Abstract: An image processing method, an image processing device, a training method of a neural network, an image processing method based on a combined neural network model, a constructing method of a combined neural network model, a neural network processor, and a storage medium are provided. The image processing method includes: obtaining, based on an input image, initial feature images of N stages with resolutions from high to low, N is a positive integer and N>2; performing, based on initial feature images of second to N-th stages, cyclic scaling processing on an initial feature image of a first stage, to obtain an intermediate feature image; and performing merging processing on the intermediate feature image to obtain an output image. The cyclic scaling processing includes hierarchically-nested scaling processing of N?1 stages, and scaling processing of each stage includes down-sampling processing, concatenating processing, up-sampling processing, and residual link addition processing.

Abstract: An undecimated image processing method includes: acquiring an image to be processed, inputting the image to be processed into an image processing network, to obtain an output image, where a resolution of the output image is the same with a resolution of the image to be processed. The inputting the image to be processed into the image processing network to obtain the output image includes: inputting the image to be processed into an analysis module to perform feature analysis, and outputting a feature tensor image; inputting the feature tensor image into a processing module, and outputting a processed feature tensor image; and synthesizing, by a synthesis module, at least one feature tensor image outputted by the at least one processing module to obtain the output image.

Abstract: An image processing method, comprising: by using a trained multi-scale detail enhancement model, performing detail enhancement on an input image to be processed; wherein multi-scale decomposition is performed on the input image to obtain a base layer image and at least one detail layer image; a first residual feature corresponding to a first feature map is acquired, and a second residual feature corresponding to a second feature map of each detail layer image is acquired; a base layer output image is obtained according to the first residual feature, each second residual feature and the first feature map, and a detail layer output image corresponding to the detail layer image is obtained according to the first residual feature, each second residual feature and the second feature map; and image fusion is performed on the base layer output image and each detail layer output image to obtain an output image.

Abstract: The present disclosure provides a video processing method, a video processing device, an electronic apparatus, and a readable storage medium. The video processing method includes: obtaining input data; and inputting the input data into a video processing model to obtain output video data. A resolution and/or a duration of the output video data is not equal to a resolution and/or a duration of the input data, the video processing model includes a plurality of generators arranged in sequence and corresponding to different image resolutions, each generator includes a transposed 3D convolution unit and a plurality of first 3D convolutional layers, the transposed 3D convolution unit is configured to generate first output data in accordance with the input data and intermediate processing data of the generator.

Abstract: An image processing method, an image processing device, a training method of a neural network, an image processing method based on a combined neural network model, a constructing method of a combined neural network model, a neural network processor, and a storage medium are provided. The image processing method includes: obtaining, based on an input image, initial feature images of N stages with resolutions from high to low, N is a positive integer and N>2; performing, based on initial feature images of second to N-th stages, cyclic scaling processing on an initial feature image of a first stage, to obtain an intermediate feature image; and performing merging processing on the intermediate feature image to obtain an output image. The cyclic scaling processing includes hierarchically-nested scaling processing of N?1 stages, and scaling processing of each stage includes down-sampling processing, concatenating processing, up-sampling processing, and residual link addition processing.

Abstract: An image processing method, an image processing device, a training method of a neural network, an image processing method based on a combined neural network model, a constructing method of a combined neural network model, a neural network processor, and a storage medium are provided. The image processing method includes: obtaining, based on an input image, initial feature images of N stages with resolutions from high to low, where N is a positive integer and N>2, performing, based on initial feature images of second to N-th stages, cyclic scaling processing on an initial feature image of a first stage, to obtain an intermediate feature image; and preforming merging processing on the intermediate feature image to obtain an output image. The cyclic scaling processing includes hierarchically-nested scaling processing of N?1 stages, and scaling processing of each stage includes down-sampling processing, concatenating processing, up-sampling processing, and residual link addition processing.

Abstract: A computer-implemented method is provided. The computer-implemented method includes rendering a dynamic effect to one or more objects in an image. The method includes setting boundary points surrounding a foreground object to define a boundary area in which a dynamic movement is to be realized; setting a movement line to define a dynamic movement direction along which the dynamic movement is to be realized, wherein setting the movement line includes detecting a continuous touch over different positions on the touch control display panel; setting a stationary area to define an area in which the dynamic movement is prohibited, wherein setting the stationary area includes detecting a first touch area corresponding to the stationary area on the touch control display panel; and generating an animation of the foreground object in the boundary area moving along the dynamic movement direction, thereby realizing the dynamic effect in the image.

Abstract: Disclosed are an image processing method and device, a training method of a neural network and a storage medium. The image processing method includes: obtaining an input image, and processing the input image by using a generative network to generate an output image. The generate network includes a first sub-network and at least one second sub-network, and the processing the input image by using the generative network to generate the output image includes, processing the input image by using the first sub-network to obtain a plurality of first feature images; performing a branching process and a weight sharing process on the plurality of first feature images by using the at least one second sub-network to obtain a plurality of second feature images; and processing the plurality of second feature images to obtain the output image.

Abstract: The disclosure provides an image reconstruction method for an edge device, an electronic device and a storage medium. The image reconstruction method includes: extracting low-level features from an input image of a first scale to generate first feature maps, the first feature maps having a second scale greater than the first scale as compared with the input image; extracting low-level features from the input image to generate second feature maps, the second feature maps having the second scale; generating mask maps based on the second feature maps; generating intermediate feature maps based on the mask maps and the first feature maps, the intermediate feature maps having the second scale; synthesizing a reconstructed image having the second scale based on the intermediate feature maps. This method facilitates to achieve a better image super-resolution reconstruction effect with lower resource consumption.

Abstract: A computer-implemented method using a convolutional neural network is provided. The computer-implemented method using a convolutional neural network includes processing an input image through at least one channel of the convolutional neural network to generate an output image including content features of the input image morphed with style features of a reference style image. The at least one channel includes a down-sampling segment, a densely connected segment, and an up-sampling segment sequentially connected together. Processing the input image through the at least one channel of the convolutional neural network includes processing an input signal through the down-sampling segment to generate a down-sampling segment output; processing the down-sampling segment output through the densely connected segment to generate a densely connected segment output; and processing the densely connected segment output through the up-sampling segment to generate an up-sampling segment output.

Abstract: A data augmentation method includes: selecting at least two different sets of samples from an original data set, each set of samples including input samples and output samples; generating at least one random number; generating at least one extended input data sample according to input samples in the at least two different sets of samples and the at least one random number; and generating at least one extended output data sample according to output samples in the at least two different sets of samples and the at least one random number, each extended input data sample corresponding to a respective extended output data sample.

Abstract: A computer-implemented method is provided. The computer-implemented method includes rendering a dynamic effect to one or more objects in an image. The method includes setting boundary points surrounding a foreground object to define a boundary area in which a dynamic movement is to be realized; setting a movement line to define a dynamic movement direction along which the dynamic movement is to be realized, wherein setting the movement line includes detecting a continuous touch over different positions on the touch control display panel; setting a stationary area to define an area in which the dynamic movement is prohibited, wherein setting the stationary area includes detecting a first touch area corresponding to the stationary area on the touch control display panel; and generating an animation of the foreground object in the boundary area moving along the dynamic movement direction, thereby realizing the dynamic effect in the image.

Abstract: A convolutional neural network, and a processing method, a processing device, a processing system and a medium for the same.

Abstract: The present disclosure generally relates to the field of deep learning technologies. An apparatus for generating a plurality of correlation images may include a feature extracting unit configured to receive a training image and extracting at least one or more of feature from the training image to generate a first feature image based on the training image; a normalizer configured to normalize the first feature image and generate a second feature image; and a shift correlating unit configured to perform a plurality of translational shifts on the second feature image to generate a plurality of shifted images, correlate each of the plurality of shifted images with the second feature image to generate the plurality of correlation images.

Abstract: Provided are a processing method and system for a convolutional neural network, and a computer-readable medium, the processing method includes training a generator and training a discriminator, wherein training a generator includes: extracting a low-resolution color image from a high-resolution color image; training parameters of a generator network, by using the low-resolution color image and a noise image as an input image, based on parameters of a discriminator network, and reducing a generator cost function; training a discriminator includes: inputting an output image of the trained generator network and the high-resolution color image to the discriminator network, respectively; training parameters of the discriminator network by reducing a discriminator cost function (S204) the generator cost function and the discriminator cost function represent a degree in which the output image of the generator network corresponds to the high-resolution color image.

Abstract: The disclosure provides an image reconstruction method for an edge device, an electronic device and a storage medium. The image reconstruction method includes: extracting low-level features from an input image of a first scale to generate first feature maps, the first feature maps having a second scale greater than the first scale as compared with the input image; extracting low-level features from the input image to generate second feature maps, the second feature maps having the second scale; generating mask maps based on the second feature maps; generating intermediate feature maps based on the mask maps and the first feature maps, the intermediate feature maps having the second scale; synthesizing a reconstructed image having the second scale based on the intermediate feature maps. This method facilitates to achieve a better image super-resolution reconstruction effect with lower resource consumption.

Abstract: A method for processing an image is provided, including: acquiring an input image; performing down-sampling and feature extraction on the input image by an encoder network to obtain multiple feature maps; and performing up-sampling and feature extraction on the multiple feature maps by a decoder network to obtain a target segmentation image; processing levels between the encoder network and the decoder network for outputting feature maps with the same resolution are connected with each other, and the encoder network and the decoder network each includes one or more dense calculation blocks, and at least one convolution module in any dense computation block includes at least one group of asymmetric convolution kernels.

Abstract: An expression recognition method is described that includes acquiring a face image to be recognized, and inputting the face image into N different recognition models arranged in sequence for expression recognition and outputting an actual expression recognition result, the N different recognition models being configured to recognize different target expression types, wherein N is an integer greater than 1.

Abstract: A training method of a neural network for implementing image style transfer, an image processing method and an image processing device are disclosed. The training method includes: acquiring a first training input image and a second training input image; performing a style transfer process on the first training input image by the neural network to obtain a training output image; based on the first training input image, the second training input image and the training output image, calculating a loss value of parameters of the neural network through a loss function; and modifying the parameters of the neural network according to the loss value, where the loss function satisfies a predetermined condition, obtaining a trained neural network, where the loss function doesn't satisfy the predetermined condition, continuing to repeatedly perform above training process, the loss function including a weight-bias-ratio loss function.