Class-Aware Fully Convolutional Gaussian and Poisson Denoising

DOI10.1109/TIP.2018.2859044MaRDI QIDQ4617539zbMATH OpenWikidataFDO

Authors Tal Remez, Or Litany, Alex M. Bronstein, Raja Giryes

Publication date 6 February 2019

Published in IEEE Transactions on Image Processing (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1808.06562

Mathematics Subject Classification ID

Neural nets and related approaches to inference from stochastic processes (62M45) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08)

Abstract: We propose a fully-convolutional neural-network architecture for image denoising which is simple yet powerful. Its structure allows to exploit the gradual nature of the denoising process, in which shallow layers handle local noise statistics, while deeper layers recover edges and enhance textures. Our method advances the state-of-the-art when trained for different noise levels and distributions (both Gaussian and Poisson). In addition, we show that making the denoiser class-aware by exploiting semantic class information boosts performance, enhances textures and reduces artifacts.

Cited in

(5)

• Deep learning on image denoising: an overview
• Shared Prior Learning of Energy-Based Models for Image Reconstruction
• Image Denoising: The Deep Learning Revolution and Beyond—A Survey Paper
• Linear and nonlinear generative models for `zero-shot' image denoising in the limit of few photons
• Proximal ADMM approach for image restoration with mixed Poisson-Gaussian noise

This page was built for publication: Class-Aware Fully Convolutional Gaussian and Poisson Denoising

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4617539)