Non-paired image generation method and system, server and storage medium
An image generation and image technology, applied in the image field, can solve the problem of poor interference suppression ability of false images, etc., and achieve the effect of enhancing differentiation and improving suppression ability
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0025] figure 1 It is a flow chart of a non-paired image generation method provided by Embodiment 1 of the present invention. This embodiment is applicable to situations where images need to be generated quickly and accurately. The method can be executed by a non-paired image generation system, which can be configured on a server superior.
[0026] Such as figure 1 As shown, the non-paired image generation method provided in the embodiment of the present invention may include:
[0027] S101. Acquire an original image, and perform instance segmentation on the original image to obtain an instance segmentation image.
[0028] Among them, the original picture is the image to be input to the image generation model. For the obtained original image, the Mask R-CNN (Mask Region-based Convolutional Neural Network) method can be used for instance segmentation. What needs to be explained here is that the instance segmentation of the original image can also use any instance segmentatio...
Embodiment 2
[0042] figure 2 It is a schematic flowchart of a method for training an image generation model provided by Embodiment 2 of the present invention. This embodiment is optimized on the basis of the above embodiments, adding an improved composition and training process of the generative confrontation network. Such as figure 2 As shown, the image generation model training method provided in the embodiment of the present invention may include:
[0043] S201. Perform instance segmentation on the training sample image to obtain an instance segmentation image based on the training sample.
[0044] S202. Fusion the training sample image and the obtained instance segmentation image.
[0045] When training the image generation model, perform instance segmentation operation and image fusion processing on the input training sample image in advance according to S201-S202 respectively, to obtain the fused training sample, wherein, the instance segmentation image based on the training sam...
Embodiment 3
[0064] image 3 It is a schematic structural diagram of an unpaired image generation system provided by Embodiment 3 of the present invention. Such as image 3 As shown, the system includes:
[0065] An instance segmentation module 301, configured to acquire an original image, and perform instance segmentation on the original image to obtain an instance segmentation image;
[0066] An instance information fusion module 302, configured to perform image fusion on the original image and the instance segmentation image;
[0067] The image generation module 303 is used to input the fused image data as an input value into the image generation model based on the improved generative confrontation network training, and obtain the target image according to the output of the image generation model, and the improved generation Adversarial networks are generative adversarial networks incorporating instance-dependent loss functions.
[0068] In the embodiment of the present invention, t...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More