Preserve More Details: Mitigating Content Drift in Real-World Image Super-Resolution

2026-08-10

AI Summary

This work addresses the challenge of content drift in real-world image super-resolution, where low-quality inputs often lead to detail degradation and semantic shifts, compromising both fidelity and perceptual quality. To mitigate this, the authors propose FSP-Diff, a novel single-step diffusion framework that introduces, for the first time, a dual-path collaborative mechanism. A detail-conditioning path injects structured fine-grained textures to restore intricate details, while a detail-modulated semantic path leverages this structural information to refine semantic guidance and suppress semantic bias. This design significantly enhances detail recovery and semantic consistency without sacrificing inference efficiency. Extensive experiments demonstrate that FSP-Diff outperforms existing single-step diffusion methods on the Real-ISR benchmark, achieving state-of-the-art performance in both quantitative metrics and visual quality.

Abstract

Real-world image super-resolution (Real-ISR) aims to reconstruct high-quality (HQ) images from low-quality (LQ) inputs subject to diverse real-world degradations. Recent advances have leveraged the LQ inputs and natural image priors learned by Stable Diffusion models to achieve impressive results. However, existing methods often overlook insufficient clarity of LQ inputs inevitably induce content drift in the generated HQ images. This manifests primarily as visual detail degradation and textual semantic shift, severely compromising both fidelity and perceptual quality. To address this challenge, we propose FSP-Diff, a novel one-step diffusion model featuring a dual-pathway architecture. This architecture comprises a Detail-Conditioned Pathway for injecting structured details to recover fine structures, and a Detail-Modulated Semantic Pathway that refines semantic guidance using structured details to mitigate semantic deviations. Extensive experiments on standard Real-ISR benchmarks demonstrate that FSP-Diff surpasses existing one-step diffusion methods in both quantitative and qualitative metrics.

Problem

• content drift
• real-world image super-resolution
• visual detail degradation
• semantic shift
• image fidelity

Innovation

• Real-ISR
• content drift
• diffusion model
• dual-pathway architecture
• structured details