Denoising single images by feature ensemble revisited

TL;DR

This paper introduces a simple, efficient image denoising architecture that uses modular concatenation to better preserve spatial fidelity and reduce smoothing, outperforming many existing models with fewer parameters.

Contribution

It revisits modular concatenation in denoising networks, demonstrating improved performance and efficiency over previous deep cascaded architectures.

Findings

Achieves significant denoising improvements over state-of-the-art methods.

Uses fewer parameters than most existing networks.

Effectively preserves spatial details and reduces smoothing artifacts.

Abstract

Image denoising is still a challenging issue in many computer vision sub-domains. Recent studies show that significant improvements are made possible in a supervised setting. However, few challenges, such as spatial fidelity and cartoon-like smoothing remain unresolved or decisively overlooked. Our study proposes a simple yet efficient architecture for the denoising problem that addresses the aforementioned issues. The proposed architecture revisits the concept of modular concatenation instead of long and deeper cascaded connections, to recover a cleaner approximation of the given image. We find that different modules can capture versatile representations, and concatenated representation creates a richer subspace for low-level image restoration. The proposed architecture's number of parameters remains smaller than the number for most of the previous networks and still achieves significant improvements over the current state-of-the-art networks.