ResMaster: Mastering High-Resolution Image Generation via Structural and Fine-Grained Guidance

TL;DR

ResMaster is a training-free method that enhances high-resolution image generation by guiding diffusion models with low-resolution references, ensuring structural coherence and fine details beyond traditional resolution limits.

Contribution

It introduces a novel patch-wise guidance technique using low-resolution references and prompts, enabling high-quality 4K image synthesis without additional training.

Findings

Sets new benchmarks for high-resolution image quality

Reduces structural distortions and pattern repetitions

Demonstrates efficiency in high-resolution generation

Abstract

Diffusion models excel at producing high-quality images; however, scaling to higher resolutions, such as 4K, often results in over-smoothed content, structural distortions, and repetitive patterns. To this end, we introduce ResMaster, a novel, training-free method that empowers resolution-limited diffusion models to generate high-quality images beyond resolution restrictions. Specifically, ResMaster leverages a low-resolution reference image created by a pre-trained diffusion model to provide structural and fine-grained guidance for crafting high-resolution images on a patch-by-patch basis. To ensure a coherent global structure, ResMaster meticulously aligns the low-frequency components of high-resolution patches with the low-resolution reference at each denoising step. For fine-grained guidance, tailored image prompts based on the low-resolution reference and enriched textual prompts produced by a vision-language model are incorporated. This approach could significantly mitigate local pattern distortions and improve detail refinement. Extensive experiments validate that ResMaster sets a new benchmark for high-resolution image generation and demonstrates promising efficiency. The project page is https://shuweis.github.io/ResMaster .