CASR: A Robust Cyclic Framework for Arbitrary Large-Scale Super-Resolution with Distribution Alignment and Self-Similarity Awareness
2026-02-25 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors address a common problem in super-resolution where making images much larger than trained for causes errors like noise and blur. They propose a method called CASR that breaks big zooms into smaller, manageable steps, using just one model to keep the image quality stable. Their technique includes special modules to keep image details aligned and restore textures, reducing mistakes that usually build up during zooming. Overall, the authors show their approach works well even when enlarging images far beyond usual limits.
Arbitrary-Scale Super-ResolutionCross-Scale Distribution ShiftCyclic Super-ResolutionSuperpixel AggregationTexture RestorationAutocorrelationSelf-SimilarityDistribution DriftImage MagnificationHigh-Frequency Textures
Authors
Wenhao Guo, Zhaoran Zhao, Peng Lu, Sheng Li, Qian Qiao, RuiDe Li
Abstract
Arbitrary-Scale SR (ASISR) remains fundamentally limited by cross-scale distribution shift: once the inference scale leaves the training range, noise, blur, and artifacts accumulate sharply. We revisit this challenge from a cross-scale distribution transition perspective and propose CASR, a simple yet highly efficient cyclic SR framework that reformulates ultra-magnification as a sequence of in-distribution scale transitions. This design ensures stable inference at arbitrary scales while requiring only a single model. CASR tackles two major bottlenecks: distribution drift across iterations and patch-wise diffusion inconsistencies. The proposed SDAM module aligns structural distributions via superpixel aggregation, preventing error accumulation, while SARM module restores high-frequency textures by enforcing autocorrelation and embedding LR self-similarity priors. Despite using only a single model, our approach significantly reduces distribution drift, preserves long-range texture consistency, and achieves superior generalization even at extreme magnification.