Abstract:Image denoising, a classic ill-posed problem, aims to recover a latent image from a noisy measurement. Over the past few decades, a considerable number of denoising methods have been studied extensively. Among these methods, supervised deep convolutional networks have garnered increasing attention, and their superior performance is attributed to their capability to learn realistic image priors from a large amount of paired noisy and clean images. However, if the image to be denoised is significantly different from the training images, it could lead to inferior results, and the networks may even produce hallucinations by using inappropriate image priors to handle an unseen noisy image. Recently, deep image prior (DIP) was proposed, and it overcame this drawback to some extent. The structure of the DIP generator network is capable of capturing the low-level statistics of a natural image using an unsupervised method with no training images other than the image itself. Compared with a supervised denoising model, the unsupervised DIP is more flexible when processing image content that must be denoised. Nevertheless, the denoising performance of DIP is usually inferior to the current supervised learning-based methods using deep convolutional networks, and it is susceptible to the over-fitting problem. To solve these problems, we propose a novel deep generative network with multiple target images and an adaptive termination condition. Specifically, we utilized mainstream denoising methods to generate two clear target images to be used with the original noisy image, enabling better guidance during the convergence process and improving the convergence speed. Moreover, we adopted the noise level estimation (NLE) technique to set a more reasonable adaptive termination condition, which can effectively solve the problem of over-fitting. Extensive experiments demonstrated that, according to the denoising results, the proposed approach significantly outperforms the original DIP method in tests on different databases. Specifically, the average peak signal-to-noise ratio (PSNR) performance of our proposed method on four databases at different noise levels is increased by 1.90 to 4.86 dB compared to the original DIP method. Moreover, our method achieves superior performance against state-of-the-art methods in terms of popular metrics, which include the structural similarity index (SSIM) and feature similarity index measurement (FSIM). Thus, the proposed method lays a good foundation for subsequent image processing tasks, such as target detection and super-resolution.

DESN: an Unsupervised MR Image Denoising Network with Deep Image Prior

Denoising of Magnetic Resonance Images with Deep Neural Regularizer Driven by Image Prior

An enhanced network for brain MR image denoising

Dilated Residual Encode-Decode Networks for Image Denoising

Denoising of 3D Magnetic Resonance Images with Multi-Channel Residual Learning of Convolutional Neural Network

Denoising of 3D Brain MR Images with Parallel Residual Learning of Convolutional Neural Network Using Global and Local Feature Extraction

PET Image Denoising Using Unsupervised Deep Learning.

Magnetic resonance image denoising for Rician noise using a novel hybrid transformer‐CNN network (HTC‐net) and self‐supervised pretraining

Global and local feature extraction based on convolutional neural network residual learning for MR image denoising

MRI Denoising Using Progressively Distribution-Based Neural Network

Laplacian Eigenmaps Network-Based Nonlocal Means Method For Mr Image Denoising

Self-supervised Deep Unrolled Reconstruction Using Regularization by Denoising

Msdra-Net: A Multi-Scale Attention-Guided Network For Magnetic Resonance Image Restoration

MR Image Denoising and Super-Resolution Using Regularized Reverse Diffusion

Image Denoising Using a Novel Deep Generative Network with Multiple Target Images and Adaptive Termination Condition

An unsupervised deep learning framework for medical image denoising

Highly Undersampled Magnetic Resonance Imaging Reconstruction Using Autoencoding Priors

Denoising of 3D Magnetic Resonance Images Using a Residual Encoder-Decoder Wasserstein Generative Adversarial Network.

HydraNet: a Multi-Branch Convolutional Neural Network Architecture for MRI Denoising

MR-self Noise2Noise: self-supervised deep learning-based image quality improvement of submillimeter resolution 3D MR images

CT Image Denoising with Perceptive Deep Neural Networks