Zoomed In, Diffused Out: Towards Local Degradation-Aware Multi-Diffusion for Extreme Image Super-Resolution
This addresses a necessary resolution limitation for image super-resolution tasks, enabling broader application of T2I models, though it is incremental as it builds on existing MultiDiffusion and prompt extraction techniques.
The paper tackles the challenge of scaling pre-trained Text-to-Image diffusion models beyond their 512x512 resolution limit for image super-resolution, enabling generation of 2K, 4K, and 8K images without additional training by using MultiDiffusion for global coherence and local degradation-aware prompts for fine structure reconstruction.
Large-scale, pre-trained Text-to-Image (T2I) diffusion models have gained significant popularity in image generation tasks and have shown unexpected potential in image Super-Resolution (SR). However, most existing T2I diffusion models are trained with a resolution limit of 512x512, making scaling beyond this resolution an unresolved but necessary challenge for image SR. In this work, we introduce a novel approach that, for the first time, enables these models to generate 2K, 4K, and even 8K images without any additional training. Our method leverages MultiDiffusion, which distributes the generation across multiple diffusion paths to ensure global coherence at larger scales, and local degradation-aware prompt extraction, which guides the T2I model to reconstruct fine local structures according to its low-resolution input. These innovations unlock higher resolutions, allowing T2I diffusion models to be applied to image SR tasks without limitation on resolution.