Method and device for constructing image editing model

Abstract

The invention discloses a method and device for constructing an image editing model, electronic equipment and a computer readable storage medium, and relates to the technical field of image processingand deep learning. The implementation scheme adopted when the image editing model is constructed comprises the steps that a training sample is acquired, wherein the training sample comprises a firstimage and a second image corresponding to the first image; constructing a generative adversarial network comprising a generator and a discriminator, wherein the generator comprises a background imagegeneration branch, a mask image generation branch and a foreground image generation branch; and training the generative adversarial network according to the first image and the second image corresponding to the first image, and taking a generator in the trained generative adversarial network as an image editing model. According to the image editing model constructed by the invention, the missing background of the generated image can be filled, so that the editing performance of the constructed image editing model is improved.

Description

technical field [0001] The present application relates to the technical field of the Internet, and in particular to a method, device, electronic device and computer-readable storage medium for constructing an image editing model in the technical fields of image processing and deep learning. Background technique [0002] There are usually two existing schemes for building image editing models: one is to model the image as a whole without distinguishing between the foreground and background of the image to generate a new image; the other is to use different models to separately image The foreground and background are modeled, and the generated foreground and background are fused to generate a new image. For the first scheme, since the foreground and background of the image are not distinguished, the constructed model cannot fill in the missing background in the generated image due to the reduction of the foreground of the input image; for the second scheme, due to the need to ...

AI Technical Summary

Problems solved by technology

For the first scheme, since the foreground and background of the image are not distinguished, the constructed model cannot fill in the missing background in the generated image due to the reduction of the foreground of the input image; for the second scheme, due to the need to use multiple Models are used to model the foreground and background of the image respectively, resulting in cumbersome model building steps and high system resource consumption.

Images