Stage GAN
Stage GANs are a class of generative adversarial networks (GANs) employing a two-stage architecture to improve image generation quality and control. Research focuses on applications ranging from 3D point cloud upsampling and continuous trajectory generation to image translation for medical anomaly detection and text-to-image synthesis. These models often leverage conditional GANs and incorporate techniques like attention mechanisms or latent keypoint control to enhance the realism and interpretability of generated outputs. The resulting advancements have significant implications for various fields, including medical imaging, urban planning, and computer graphics.
Papers
May 22, 2023
January 16, 2023
July 21, 2022
July 6, 2022
May 6, 2022