Low-Res Leads the Way: Improving Generalization for Super-Resolution by Self-Supervised Learning

For image super-resolution (SR), bridging the gap between the performance onsynthetic datasets and real-world degradation scenarios remains a challenge.This work introduces a novel "Low-Res Leads the Way" (LWay) training framework,merging Supervised Pre-training with Self-supervised Learning to enhance theadaptability of SR models to real-world images. Our approach utilizes alow-resolution (LR) reconstruction network to extract degradation embeddingsfrom LR images, merging them with super-resolved outputs for LR reconstruction.Leveraging unseen LR images for self-supervised learning guides the model toadapt its modeling space to the target domain, facilitating fine-tuning of SRmodels without requiring paired high-resolution (HR) images. The integration ofDiscrete Wavelet Transform (DWT) further refines the focus on high-frequencydetails. Extensive evaluations show that our method significantly improves thegeneralization and detail restoration capabilities of SR models on unseenreal-world datasets, outperforming existing methods. Our training regime isuniversally compatible, requiring no network architecture modifications, makingit a practical solution for real-world SR applications.