Tencent Technology (Shenzhen) Company Limited

Tsinghua University

Tencent Technology Company Limited

Shimin Hu

Jianyu Wang

Haifeng Deng

Baoli Li

Xianying Li

Yan Zhang

Feiyue Huang

Bin Hua Liu

Miao Wang

Juhong Wang

Xiaowei Guo

Bin Liu

Haozhi Huang

Kang Chen

Ruixin Zhang

Kun Xu

Fanglue Zhang

Yukun Lai

Senzhe Xu

Songpei Du

Masia Belen

Gutierrez Diego

Weilun Sun

Methods, devices, and storage medium for fusing at least one image are disclosed. The method includes obtaining a first to-be-fused image and a second to-be-fused image, the first to-be-fused image comprising first regions, and the second to-be-fused image comprising second regions; obtaining a first feature set according to the first to-be-fused image and obtaining a second feature set according to the second to-be-fused image; performing first fusion processing on the first to-be-fused image and the second to-be-fused image by using a shape fusion network model to obtain a third to-be-fused image, the third to-be-fused image comprising at least one first encoding feature and at least one second encoding feature; and performing second fusion processing on the third to-be-fused image and the first to-be-fused image by using a condition fusion network model to obtain a target fused image. Model training methods, apparatus, and storage medium are also disclosed.

Image fusion method, model training method, and related apparatuses

An image classification method is provided. The method includes: inputting a to-be-classified image into a plurality of neural network models; obtaining data output by multiple non-input layers specified by each neural network model to generate a plurality of image features corresponding to the plurality of neural network models; respectively inputting the plurality of corresponding image features into linear classifiers, each of the linear classifiers being trained by one of the plurality of neural network models for determining whether an image belongs to a preset class; obtaining, using each neural network model, a corresponding probability that the to-be-classified image comprises an object image of the preset class; and determining, according to each obtained probability, whether the to-be-classified image includes the object image of the preset class.

Image classification method, electronic device, and storage medium

The present invention discloses a method for acquiring a comfort degree of a motion-sensing binocular stereoscopic video. The method includes the following steps: S. carrying out binocular parallax and movement speed analysis on an input binocular video, and calculating a binocular parallax d, a movement speed vxy in a visual plane direction and a movement speed vz in a depth direction of each pixel point in the video; S. calculating a corresponding comfort degree value c according to the parameters (d, vxy, vz) of the pixel point; S. calculating the comfort degree value of the entire frame according to the calculated comfort degree value of each pixel point; S. calculating the comfort degree value of the entire video according to the comfort degree value of each frame. The comfort degree of the binocular stereoscopic video is calculated according to the steps in the present invention, so that influence of positive and negative depth ranges and object movement on the comfort degree can be well considered to obtain more accurate estimation of the comfort degree.

Method for acquiring comfort degree of motion-sensing binocular stereoscopic video

Provided are a method and an apparatus for adjusting image brightness. The method includes: acquiring a single-channel brightness image based on grayscales of each channel of an image to be processed; performing Gaussian filtering on the single-channel brightness image to acquire a Gaussian filtered image; adjusting grayscales of the Gaussian filtered image based on the grayscales of the Gaussian filtered image and a preset proportion; acquiring a grayscale change rate of each pixel with respect to the Gaussian filtered image after adjustment and the Gaussian filtered image before adjustment; and processing the image to be processed based on the grayscale change rate of each pixel to acquire a processed image.

Method and apparatus for adjusting image brightness

The present invention relates to the technical field of image processing, and in particular, to a patch net model for image representation and a construction method of the patch net model. The patch net model for image representation is of a forest-shaped structure consisting of a plurality of composite nodes and basic nodes, each composite node is a non-leaf node, and each basic node is a leaf node; the basic node includes a certain patch region of an image and a representative patch representing an apparent feature of the patch region; the composite node includes a certain patch region of the image and can be further decomposed into basic nodes and/or composite nodes; an edge exists between two nodes, which are located on the same layer of the forest-shaped structure and are spatially connected, and a relation matrix used for expressing the spatial relative position of the two nodes is arranged on the edge.

Patch net model and construction method thereof

Provided are a method and an apparatus for adjusting image brightness. The method includes: acquiring an image to be processed, and acquiring a single-channel brightness image based on grayscales of each channel of the image to be processed; performing Gaussian filtering on the single-channel brightness image to acquire a Gaussian filtered image; adjusting grayscales of the Gaussian filtered image based on the grayscales of the Gaussian filtered image and a preset proportion; comparing the Gaussian filtered image after adjustment with the Gaussian filtered image before adjustment to acquire a grayscale change rate of each pixel through adjustment; processing the image to be processed based on the grayscale change rate of each pixel to acquire a processed image; and outputting the processed image.

The disclosure relates to an image tone adjustment method, an apparatus thereof, and a computer-readable storage medium. The method includes: fitting overlapping areas to obtain a first correction parameter and a second correction parameter, performing interpolation on each pixel of the target image to obtain a corresponding correction parameter, correcting the target image according to the corresponding correction parameters; fitting an overlapping area between the corrected target image and the first image, and an overlapping area between the corrected target image and the second image to obtain a first RGB three-channel linear transformation parameter and a second RGB three-channel linear transformation parameter, performing linear interpolation on each pixel of the corrected target image to obtain a corresponding RGB three-channel linear transformation parameter, and transforming the corrected target image.

Image tone adjustment method, apparatus thereof and computer storage medium

The present invention discloses a method for rapidly vectorizing an image by gradient meshes based on parameterization, which comprises the following steps: determining an image region to be vectorized (S); converting the image region into a mesh representation (S); mapping the meshes to a planar rectangular region by parameterizing the meshes (S); and generating a gradient mesh image according to the parameterization result of said meshes (S). The present invention generates gradient meshes by converting an image region into meshes and by parameterization, so the gradient meshes are obtained completely automatically without the need for the user to give original meshes and moreover, the computation speed is improved significantly since nonlinear optimization is avoided. In addition, the method of the present invention can process image regions containing or not containing holes.

Method and system for rapidly vectorizing image by gradient meshes based on parameterization

Growing community of inventors

Beijing, China

Shimin Hu