AI summaryⓘ
The authors address the problem of hallucination in generative super-resolution, where the generated high-resolution images contain details not well supported by the low-resolution input. They note that existing methods mix preserving coarse, reliable information and adding fine details without explicitly separating these steps. By using a visual autoregressive model, they propose K2N, which starts with the coarse image directly from the input and only generates finer details autoregressively, reducing error accumulation. Their experiments show that K2N performs well on standard quality measures and better controls hallucinated details. This suggests that clearly dividing the generation process by scale can improve super-resolution reliability.
Generative super-resolutionHallucinationVisual autoregressive modelingCoarse-to-fine generationScale-wise generationLow-resolution inputDetail continuationReconstruction fidelityPerceptual realismAutoregressive generation
Authors
Hongyi Fang, Jiahui Wu, Yichen Yue, Benjia Zhou, Dan Zeng
Abstract
Hallucination remains a persistent challenge in generative super-resolution (GSR), where reconstructed results may contain visually plausible yet weakly supported content, structural deviations, or unnatural textures with respect to the low-resolution (LR) input. Existing GSR methods have extensively explored the trade-off between perceptual realism and reconstruction fidelity, but the division between preserving reliable coarse-scale information and restoring more uncertain fine details is often handled implicitly within the overall restoration process. Visual autoregressive (VAR) modeling provides a natural opportunity to revisit this issue, as its coarse-to-fine next-scale prediction offers an explicit scale-wise generation interface. However, existing VAR-based SR methods still inherit the original full 1-to-$N$ autoregressive generation path, even though, for super-resolution, coarse-scale information in LR is often relatively more reliable, while long autoregressive chains may accumulate prediction errors. Motivated by these observations, we propose \textbf{K2N}, which reformulates VAR-based SR from full-path generation into a $k$-to-$N$ detail continuation process. Specifically, early coarse-scale states are established directly from LR, while only the remaining finer scales are restored autoregressively. Experimental results show that K2N remains competitive with the VARSR baseline on standard SR metrics, while exhibiting clearer advantages on hallucination-focused evaluation. These findings suggest that explicitly rethinking the generation path in a scale-wise manner can be a promising direction for improving the reliability of generative super-resolution. Our code will be released soon at https://github.com/BRL-SYSU/K2NSR.