Repository files navigation

Kai Zhang, Yawei Li, Wangmeng Zuo, Lei Zhang, Luc Van Gool, Radu Timofte

Computer Vision Lab, ETH Zurich, Switzerland

[paper arxiv] [paper tpami]

PSNR(Y) means the PSNR is calculated on the Y channel of YCbCr space.

Recent works on plug-and-play image restoration have shown that a denoiser can implicitly serve as the image prior formodel-based methods to solve many inverse problems. Such a property induces considerable advantages for plug-and-play imagerestoration (e.g., integrating the flexibility of model-based method and effectiveness of learning-based methods) when the denoiser isdiscriminatively learned via deep convolutional neural network (CNN) with large modeling capacity. However, while deeper and largerCNN models are rapidly gaining popularity, existing plug-and-play image restoration hinders its performance due to the lack of suitabledenoiser prior. In order to push the limits of plug-and-play image restoration, we set up a benchmark deep denoiser prior by training ahighly flexible and effective CNN denoiser. We then plug the deep denoiser prior as a modular part into a half quadratic splitting basediterative algorithm to solve various image restoration problems. We, meanwhile, provide a thorough analysis of parameter setting,intermediate results and empirical convergence to better understand the working mechanism. Experimental results on threerepresentative image restoration tasks, including deblurring, super-resolution and demosaicing, demonstrate that the proposedplug-and-play image restoration with deep denoiser prior not only significantly outperforms other state-of-the-art model-based methodsbut also achieves competitive or even superior performance against state-of-the-art learning-based methods.

• Network architecture

  

• Grayscale image denoising

  

• Color image denoising

  

(a) Noisy image with noise level 200	(b) Result by the proposed DRUNet denoiser

Even trained on noise level range of [0, 50], DRUNet can still perform well on an extremely large unseen noise level of 200.

• Visual comparison

  

• Visual comparison

  

• PSNR

  

• Visual comparison

  

@article{zhang2021plug,title={Plug-and-Play Image Restoration with Deep Denoiser Prior},author={Zhang, Kai and Li, Yawei and Zuo, Wangmeng and Zhang, Lei and Van Gool, Luc and Timofte, Radu},journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},volume={44},number={10},pages={6360-6376},year={2021}}@inproceedings{zhang2017learning,title={Learning Deep CNN Denoiser Prior for Image Restoration},author={Zhang, Kai and Zuo, Wangmeng and Gu, Shuhang and Zhang, Lei},booktitle={IEEE Conference on Computer Vision and Pattern Recognition},pages={3929--3938},year={2017}, }