Saved in:
Bibliographic Details
Main Authors: Min, Chen, Jiang, Enze, Peng, Jishen, Ma, Zheng
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.02193
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • Inverting real images into the noise space is essential for editing tasks using diffusion models, yet existing methods produce non-Gaussian noise with poor editability due to the inaccuracy in early noising steps. We identify the root cause: a mathematical singularity that renders inversion fundamentally ill-posed. We propose Singularity Skipping Inversion of Diffusion Models (SSI-DM), which bypasses this singular region by adding small noise before standard inversion. This simple approach produces inverted noise with natural Gaussian properties while maintaining reconstruction fidelity. As a plug-and-play technique compatible with general diffusion models, our method achieves superior performance on public image datasets for reconstruction and interpolation tasks, providing a principled and efficient solution to diffusion model inversion.