FiDeSR: High-Fidelity and Detail-Preserving One-Step Diffusion Super-Resolution
FiDeSR addresses the problem of image super-resolution for computer vision applications, providing an incremental improvement over existing diffusion-based methods.
FiDeSR tackles the problem of image super-resolution by achieving high-fidelity and detail-preserving reconstruction, resulting in superior real-world performance compared to existing diffusion-based methods. The framework produces outputs with both high perceptual quality and faithful content restoration.
Diffusion-based approaches have recently driven remarkable progress in real-world image super-resolution (SR). However, existing methods still struggle to simultaneously preserve fine details and ensure high-fidelity reconstruction, often resulting in suboptimal visual quality. In this paper, we propose FiDeSR, a high-fidelity and detail-preserving one-step diffusion super-resolution framework. During training, we introduce a detail-aware weighting strategy that adaptively emphasizes regions where the model exhibits higher prediction errors. During inference, low- and high-frequency adaptive enhancers further refine the reconstruction without requiring model retraining, enabling flexible enhancement control. To further improve the reconstruction accuracy, FiDeSR incorporates a residual-in-residual noise refinement, which corrects prediction errors in the diffusion noise and enhances fine detail recovery. FiDeSR achieves superior real-world SR performance compared to existing diffusion-based methods, producing outputs with both high perceptual quality and faithful content restoration. The source code will be released at: https://github.com/Ar0Kim/FiDeSR.