GuideSR: Rethinking Guidance for One-Step High-Fidelity Diffusion-Based Super-Resolution

TL;DR

GuideSR introduces a dual-branch diffusion model that significantly improves image super-resolution fidelity and perceptual quality by preserving structural details and leveraging pre-trained diffusion models.

Contribution

The paper presents a novel dual-branch architecture for single-step diffusion-based super-resolution, effectively combining structural preservation with perceptual enhancement.

Findings

Achieves up to 1.39dB PSNR gain on real-world datasets.

Outperforms existing methods on multiple reference-based metrics.

Maintains low computational cost of single-step approaches.

Abstract

In this paper, we propose GuideSR, a novel single-step diffusion-based image super-resolution (SR) model specifically designed to enhance image fidelity. Existing diffusion-based SR approaches typically adapt pre-trained generative models to image restoration tasks by adding extra conditioning on a VAE-downsampled representation of the degraded input, which often compromises structural fidelity. GuideSR addresses this limitation by introducing a dual-branch architecture comprising: (1) a Guidance Branch that preserves high-fidelity structures from the original-resolution degraded input, and (2) a Diffusion Branch, which a pre-trained latent diffusion model to enhance perceptual quality. Unlike conventional conditioning mechanisms, our Guidance Branch features a tailored structure for image restoration tasks, combining Full Resolution Blocks (FRBs) with channel attention and an Image Guidance Network (IGN) with guided attention. By embedding detailed structural information directly into the restoration pipeline, GuideSR produces sharper and more visually consistent results. Extensive experiments on benchmark datasets demonstrate that GuideSR achieves state-of-the-art performance while maintaining the low computational cost of single-step approaches, with up to 1.39dB PSNR gain on challenging real-world datasets. Our approach consistently outperforms existing methods across various reference-based metrics including PSNR, SSIM, LPIPS, DISTS and FID, further representing a practical advancement for real-world image restoration.