Image and Texture Independent Deep Learning Noise Estimation using Multiple Frames

TL;DR

This paper introduces a novel CNN-based noise estimation method that utilizes multiple frames to achieve image and texture independence, improving noise estimation accuracy across various conditions.

Contribution

The paper presents a new multi-frame CNN noise estimator that is independent of image content and texture, advancing noise estimation techniques.

Findings

Effective noise estimation across diverse textures

Improved accuracy over single-frame methods

Robustness to different image conditions

Abstract

In this study, a novel multiple-frame based image and texture independent convolutional Neural Network (CNN) noise estimator is introduced. The estimator works.