EditCrafter: Tuning-free High-Resolution Image Editing via Pretrained Diffusion Model

TL;DR

EditCrafter is a tuning-free high-resolution image editing method that leverages pretrained diffusion models, enabling high-quality editing at resolutions beyond training limits without fine-tuning.

Contribution

It introduces a novel pipeline with tiled inversion and noise-damped guidance, allowing high-resolution editing without tuning or optimization.

Findings

Achieves high-quality editing across various resolutions.

Operates without fine-tuning or optimization.

Handles arbitrary aspect ratios effectively.

Abstract

We propose EditCrafter, a high-resolution image editing method that operates without tuning, leveraging pretrained text-to-image (T2I) diffusion models to process images at resolutions significantly exceeding those used during training. Leveraging the generative priors of large-scale T2I diffusion models enables the development of a wide array of novel generation and editing applications. Although numerous image editing methods have been proposed based on diffusion models and exhibit high-quality editing results, they are difficult to apply to images with arbitrary aspect ratios or higher resolutions since they only work at the training resolutions (512x512 or 1024x1024). Naively applying patch-wise editing fails with unrealistic object structures and repetition. To address these challenges, we introduce EditCrafter, a simple yet effective editing pipeline. EditCrafter operates by first performing tiled inversion, which preserves the original identity of the input high-resolution image. We further propose a noise-damped manifold-constrained classifier-free guidance (NDCFG++) that is tailored for high resolution image editing from the inverted latent. Our experiments show that the our EditCrafter can achieve impressive editing results across various resolutions without fine-tuning and optimization.