Dual-Schedule Inversion: Training- and Tuning-Free Inversion for Real Image Editing

TL;DR

This paper introduces Dual-Schedule Inversion, a novel method for real image editing that achieves perfect reconstruction without fine-tuning and ensures semantic consistency in edited images.

Contribution

It proposes a new inversion and sampling technique that overcomes DDIM Inversion limitations, enabling high-quality, fine-tuning-free real image editing with guaranteed reversibility.

Findings

Achieves perfect reconstruction of real images without fine-tuning

Ensures edited images conform to text prompt semantics

Retains original identity in unedited parts

Abstract

Text-conditional image editing is a practical AIGC task that has recently emerged with great commercial and academic value. For real image editing, most diffusion model-based methods use DDIM Inversion as the first stage before editing. However, DDIM Inversion often results in reconstruction failure, leading to unsatisfactory performance for downstream editing. To address this problem, we first analyze why the reconstruction via DDIM Inversion fails. We then propose a new inversion and sampling method named Dual-Schedule Inversion. We also design a classifier to adaptively combine Dual-Schedule Inversion with different editing methods for user-friendly image editing. Our work can achieve superior reconstruction and editing performance with the following advantages: 1) It can reconstruct real images perfectly without fine-tuning, and its reversibility is guaranteed mathematically. 2) The edited object/scene conforms to the semantics of the text prompt. 3) The unedited parts of the object/scene retain the original identity.