Implicit Diffusion Models for Continuous Super-Resolution

Image super-resolution (SR) has attracted increasing attention due to its wide applications. However, current SR methods generally suffer from over-smoothing and artifacts, and most work only with fixed magnifications. This paper introduces an Implicit Diffusion Model (IDM) for high-fidelity continuous image super-resolution. IDM integrates an implicit neural representation and a denoising diffusion model in a unified end-to-end framework, where the implicit neural representation is adopted in the decoding process to learn continuous-resolution representation. Furthermore, we design a scale-controllable conditioning mechanism that consists of a low-resolution (LR) conditioning network and a scaling factor. The scaling factor regulates the resolution and accordingly modulates the proportion of the LR information and generated features in the final output, which enables the model to accommodate the continuous-resolution requirement. Extensive experiments validate the effectiveness of our IDM and demonstrate its superior performance over prior arts.

PDF Abstract CVPR 2023 PDF CVPR 2023 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Result Benchmark
Image Super-Resolution CelebA-HQ 128x128 SR3 PSNR 23.04 # 2
SSIM 0.65 # 2
Image Super-Resolution CelebA-HQ 128x128 IDM PSNR 24.01 # 1
SSIM 0.71 # 1
Image Super-Resolution DIV2K val - 4x upscaling LAR-SR PSNR 27.03 # 14
SSIM 0.77 # 12
Image Super-Resolution DIV2K val - 4x upscaling HCFlow++ PSNR 26.61 # 18
SSIM 0.74 # 20
Image Super-Resolution DIV2K val - 4x upscaling HCFlow PSNR 27.02 # 15
SSIM 0.76 # 14
Image Super-Resolution DIV2K val - 4x upscaling RankSRGAN PSNR 26.55 # 19
SSIM 0.75 # 17
Image Super-Resolution DIV2K val - 4x upscaling ESRGAN PSNR 26.22 # 20
SSIM 0.75 # 17
Image Super-Resolution DIV2K val - 4x upscaling Bicubic PSNR 26.7 # 17
SSIM 0.77 # 12
Image Super-Resolution DIV2K val - 4x upscaling IDM PSNR 27.59 # 9
SSIM 0.78 # 9

Methods