This project used Graphical Generative Adversarial Networks.Using a combination of graph structure and depth implicit model,it can simulate the trajectory of targets in a series of images.So we ues it to generate new sar images based on existed sar images,so that if we can use generated sar images to better classify targets.To do that,we modified ssgan_inference_chairs.py to generate new sar images.
ghailey / sar_image_generate Goto Github PK
View Code? Open in Web Editor NEWUse Graphical Generative Adversarial Networks to generate new sar images
License: MIT License