Image Super-Resolution via Iterative Refinement

Abstract

We present SR3, an approach to image Super-Resolution via Repeated Refinement. SR3 adapts denoising diffusion probabilistic models to conditional image generation and performs super-resolution through a stochastic denoising process. Inference starts with pure Gaussian noise and iteratively refines the noisy output using a U-Net model trained on denoising at various noise levels. SR3 exhibits strong performance on super-resolution tasks at different magnification factors, on faces and natural images. We conduct human evaluation on a standard 8X face super-resolution task on CelebA-HQ, comparing with SOTA GAN methods. SR3 achieves a fool rate close to 50%, suggesting photo-realistic outputs, while GANs do not exceed a fool rate of 34%. We further show the effectiveness of SR3 in cascaded image generation, where generative models are chained with super-resolution models, yielding a competitive FID score of 11.3 on ImageNet.

Cited by in corpus (5)

• Break-A-Scene: Extracting Multiple Concepts from a Single Image
• 4D Facial Expression Diffusion Model
• Deblending Galaxies with Generative Adversarial Networks
• Conditional Distribution Modelling for Few-Shot Image Synthesis with Diffusion Models
• Stochastic Super-resolution of Cosmological Simulations with Denoising Diffusion Models