Gated Texture CNN for Efficient and Configurable Image Denoising

TL;DR

The paper introduces Gated Texture CNN (GTCNN), a novel image denoising model that efficiently preserves textures and allows interactive control over texture strength, outperforming previous methods with fewer parameters.

Contribution

GTCNN is the first CNN-based denoising model that effectively excludes texture information from intermediate features using gating, enabling efficient and controllable denoising.

Findings

Achieves state-of-the-art denoising performance.

Uses 4.8 times fewer parameters than previous models.

Allows interactive texture control without extra costs.

Abstract

Convolutional neural network (CNN)-based image denoising methods typically estimate the noise component contained in a noisy input image and restore a clean image by subtracting the estimated noise from the input. However, previous denoising methods tend to remove high-frequency information (e.g., textures) from the input. It caused by intermediate feature maps of CNN contains texture information. A straightforward approach to this problem is stacking numerous layers, which leads to a high computational cost. To achieve high performance and computational efficiency, we propose a gated texture CNN (GTCNN), which is designed to carefully exclude the texture information from each intermediate feature map of the CNN by incorporating gating mechanisms. Our GTCNN achieves state-of-the-art performance with 4.8 times fewer parameters than previous state-of-the-art methods. Furthermore, the GTCNN allows us to interactively control the texture strength in the output image without any additional modules, training, or computational costs.