Abstract:Real-world image noise removal is a long-standing yet very challenging task in computer vision. The success of deep neural network in denoising stimulates the research of noise generation, aiming at synthesizing more clean-noisy image pairs to facilitate the training of deep denoisers. In this work, we propose a novel unified framework to simultaneously deal with the noise removal and noise generation tasks. Instead of only inferring the posteriori distribution of the latent clean image conditioned on the observed noisy image in traditional MAP framework, our proposed method learns the joint distribution of the clean-noisy image pairs. Specifically, we approximate the joint distribution with two different factorized forms, which can be formulated as a denoiser mapping the noisy image to the clean one and a generator mapping the clean image to the noisy one. The learned joint distribution implicitly contains all the information between the noisy and clean images, avoiding the necessity of manually designing the image priors and noise assumptions as traditional. Besides, the performance of our denoiser can be further improved by augmenting the original training dataset with the learned generator. Moreover, we propose two metrics to assess the quality of the generated noisy image, for which, to the best of our knowledge, such metrics are firstly proposed along this research line. Extensive experiments have been conducted to demonstrate the superiority of our method over the state-of-the-arts both in the real noise removal and generation tasks. The training and testing code is available at https://github.com/zsyOAOA/DANet.

Adversarial Signal Denoising with Encoder-Decoder Networks

Image denoising algorithm based on adversarial learning using joint loss function

Denoising Adversarial Autoencoders

LDN-RC: a Lightweight Denoising Network with Residual Connection to Improve Adversarial Robustness

Removing Noise from Extracellular Neural Recordings Using Fully Convolutional Denoising Autoencoders

Dual Adversarial Network: Toward Real-world Noise Removal and Noise Generation

Deep Recurrent Neural Networks for ECG Signal Denoising

Image Denoising Using Very Deep Fully Convolutional Encoder-Decoder Networks with Symmetric Skip Connections.

Eliminating Adversarial Noise Via Information Discard and Robust Representation Restoration.

Analyzing noise in autoencoders and deep networks

Wireless Signal Denoising Using Conditional Generative Adversarial Networks.

Adversarial Distortion Learning for Medical Image Denoising

Learning from Natural Noise to Denoise Micro-Doppler Spectrogram

Learning Model-Blind Temporal Denoisers without Ground Truths

Robust Vibration Signal Denoising and Diagnosis Using Encoder-Decoder Networks with Cross-layer Residual Connection

Deep Generative Adversarial Networks for the Sparse Signal Denoising

Denoising: from classical methods to deep CNNs

A signal processing interpretation of noise-reduction convolutional neural networks

Denoising Noisy Neural Networks: A Bayesian Approach with Compensation.

Environmental Noise Reduction based on Deep Denoising Autoencoder

Denoising Auto-Encoders Toward Robust Unsupervised Feature Representation.