ScrollScape: Unlocking 32K Image Generation With Video Diffusion Priors

TL;DR

ScrollScape transforms ultra-high-resolution image synthesis into a video generation problem, leveraging video priors to maintain structural integrity at 32K resolution with extreme aspect ratios.

Contribution

It introduces a novel framework that reformulates EAR image synthesis as video generation, utilizing spatial-temporal mapping and super-resolution priors for unprecedented scale and quality.

Findings

Outperforms existing baselines by reducing artifacts.

Achieves 32K resolution with global coherence.

Effectively aligns video priors with high-resolution image synthesis.

Abstract

While diffusion models excel at generating images with conventional dimensions, pushing them to synthesize ultra-high-resolution imagery at extreme aspect ratios (EAR) often triggers catastrophic structural failures, such as object repetition and spatial fragmentation. This limitation fundamentally stems from a lack of robust spatial priors, as static text-to-image models are primarily trained on image distributions with conventional dimensions. To overcome this bottleneck, we present ScrollScape, a novel framework that reformulates EAR image synthesis into a continuous video generation process through two core innovations. By mapping the spatial expansion of a massive canvas to the temporal evolution of video frames, ScrollScape leverages the inherent temporal consistency of video models as a powerful global constraint to ensure long-range structural integrity. Specifically, Scanning Positional Encoding (ScanPE) distributes global coordinates across frames to act as a flexible moving camera, while Scrolling Super-Resolution (ScrollSR) leverages video super-resolution priors to circumvent memory bottlenecks, efficiently scaling outputs to an unprecedented 32K resolution. Fine-tuned on a curated 3K multi-ratio image dataset, ScrollScape effectively aligns pre-trained video priors with the EAR generation task. Extensive evaluations demonstrate that it significantly outperforms existing image-diffusion baselines by eliminating severe localized artifacts. Consequently, our method overcomes inherent structural bottlenecks to ensure exceptional global coherence and visual fidelity across diverse domains at extreme scales.