Image Restoration with Mean-Reverting Stochastic Differential Equations

By Ziwei Luo, Fredrik K. Gustafsson, Zheng Zhao, Jens Sjölund, Thomas Bo Schön. International Conference on Machine Learning, 2023

318 citations recorded by the source corpus.

This paper presents a stochastic differential equation (SDE) approach for general-purpose image restoration. The key construction consists in a mean-reverting SDE that transforms a high-quality image into a degraded counterpart as a mean state with fixed Gaussian noise. Then, by simulating the corresponding reverse-time SDE, we are able to restore the origin of the low-quality image without relying on any task-specific prior knowledge. Crucially, the proposed mean-reverting SDE has a closed-form solution, allowing us to compute the ground truth time-dependent score and learn it with a neural network. Moreover, we propose a maximum likelihood objective to learn an optimal reverse trajectory that stabilizes the training and improves the restoration results. The experiments show that our proposed method achieves highly competitive performance in quantitative comparisons on image deraining, deblurring, and denoising, setting a new state-of-the-art on two deraining datasets. Finally, the general applicability of our approach is further demonstrated via qualitative results on image super-resolution, inpainting, and dehazing. Code is available at https://github.com/Algolzw/image-restoration-sde.

Subfields: Biophysics, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/c9217c50acce84e417bbc13a4ecaa06db0c77026

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biophysics

  1. Chen Change Loy
  2. Jianyi Wang

In Computer Vision and Pattern Recognition

  1. Zongsheng Yue
  2. Jianyi Wang
  3. Chen Change Loy
  4. Lei Zhang
  5. Yu Qiao
  6. Luc Van Gool
  7. Chao Dong
  8. Xiaogang Wang

Citations in the corpus