Denoising Prior Driven Deep Neural Network for Image Restoration

By Weisheng Dong, Peiyao Wang, Wotao Yin, Guangming Shi, Fangfang Wu, Xiaotong Lu. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018

491 citations recorded by the source corpus.

Deep neural networks (DNNs) have shown very promising results for various image restoration (IR) tasks. However, the design of network architectures remains a major challenging for achieving further improvements. While most existing DNN-based methods solve the IR problems by directly mapping low quality images to desirable high-quality images, the observation models characterizing the image degradation processes have been largely ignored. In this paper, we first propose a denoising-based IR algorithm, whose iterative steps can be computed efficiently. Then, the iterative process is unfolded into a deep neural network, which is composed of multiple denoisers modules interleaved with back-projection (BP) modules that ensure the observation consistencies. A convolutional neural network (CNN) based denoiser that can exploit the multi-scale redundancies of natural images is proposed. As such, the proposed network not only exploits the powerful denoising ability of DNNs, but also leverages the prior of the observation model. Through end-to-end training, both the denoisers and the BP modules can be jointly optimized. Experimental results on several IR tasks, e.g., image denoisig, super-resolution and deblurring show that the proposed method can lead to very competitive and often state-of-the-art results on several IR tasks, including image denoising, deblurring, and super-resolution.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/8bac6c54aaaa53c8573508efddc0425e833e0560

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Image denoising using deep CNN with batch renormalization (2020, methodology): “A CNN with a prior was also very effective for noise of certain type [10, 66, 13].”
Residual Dense Network for Image Restoration (2018, methodology): “A common practice to synthesize blurry images is first apply a blur kernel and then add Gausian noise with noise level s [7], [8].”
Learning to Optimize: A Primer and A Benchmark (2021, unknown): “Some other PnP schemes were analyzed under different assumptions [203; 204; 35; 65; 207; 39; 197; 71; 238; 199; 198].”
Image De-Raining Transformer (2022, unknown): “In recent years, deep learning are widely used in various image processing tasks [17], [18], [31], [32], [33], [34], [35], such as de-noising [36], [37], [38], super-resolution [39], [40], [41], de-blurring [42], [43], [44], [45], de-hazing [19], [19], XIAO ETAL.”
Deep Generalized Unfolding Networks for Image Restoration (2022, methodology): “For instance, [16] jointly optimized an UNet as the proximal mapping in the ADMM [5] algorithm.”
Plug-and-Play Methods Provably Converge with Properly Trained Denoisers (2019, methodology): “…ptychographic microscopy (Sun et al., 2018b), low-dose CT imaging (Venkatakrishnan et al., 2013; He et al., 2018; Ye et al., 2018; Lyu et al., 2019), hyper-spectral sharpening (Teodoro et al., 2017; 2019), inpainting (Chan, 2019; Tirer & Giryes, 2019), and superresolution (Dong et al., 2018).”
AMP-Net: Denoising-Based Deep Unfolding for Compressive Image Sensing (2020, methodology): “rained with other parameters using gradient descend related algorithms [45]. Since the sampling matrix plays an important role in both sampling and reconstruction in most deep unfolding methods [34], [36], [41], an appropriate sampling matrix may effectively improve the reconstruction performance. I”