Image Super-Resolution via Iterative Refinement
This provides a novel method for generating photo-realistic high-resolution images, addressing a key problem in computer vision for applications like media enhancement.
The paper tackles image super-resolution by adapting denoising diffusion models to generate high-resolution images from low-resolution inputs, achieving a fool rate close to 50% on face super-resolution and a competitive FID score of 11.3 on ImageNet.
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.