Northeastern University and Microsoft Researchers Propose A Novel Two-Branch Technique That Expands StyleGAN’s Latent Space

Generative adversarial networks (GANs) have been widely utilized for image synthesis. In 2019, NVIDIA open-sourced Style Generative Adversarial Network, or StyleGAN for short, is a GAN architecture extension that modifies the generator model significantly. StyleGAN has capabilities to create the simulated image in steps, starting with a low resolution and increasing to a high resolution. The architecture achieves this by evaluating the visual features that are manifested in that level, from standard aspects (position, face shape) to minute details (hair …

computervision microsoft researchers space university