Reversible Inversion for Training-Free Exemplar-guided Image Editing

2025-12-01

AI Summary

Existing exemplar-guided image editing methods either rely on costly pre-trained models or adopt training-free inversion strategies that suffer from poor editing quality and low efficiency. This paper proposes ReInversion—a training-free, invertible inversion framework that achieves precise latent-space alignment between source and reference images via two-stage conditional denoising. It further introduces a mask-guided selective denoising mechanism to ensure accurate editing region localization while preserving background structural integrity. To the best of our knowledge, ReInversion is the first method enabling high-quality, efficient exemplar-guided editing entirely without training or fine-tuning. It achieves state-of-the-art editing fidelity across multiple benchmarks while significantly reducing computational overhead and maintaining high reconstruction fidelity.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Exemplar-guided Image Editing (EIE) aims to modify a source image according to a visual reference. Existing approaches often require large-scale pre-training to learn relationships between the source and reference images, incurring high computational costs. As a training-free alternative, inversion techniques can be used to map the source image into a latent space for manipulation. However, our empirical study reveals that standard inversion is sub-optimal for EIE, leading to poor quality and inefficiency. To tackle this challenge, we introduce extbf{Reversible Inversion ({ReInversion})} for effective and efficient EIE. Specifically, ReInversion operates as a two-stage denoising process, which is first conditioned on the source image and subsequently on the reference. Besides, we introduce a Mask-Guided Selective Denoising (MSD) strategy to constrain edits to target regions, preserving the structural consistency of the background. Both qualitative and quantitative comparisons demonstrate that our ReInversion method achieves state-of-the-art EIE performance with the lowest computational overhead.

Problem

• Improves training-free exemplar-guided image editing efficiency
• Enhances inversion quality for source-reference image manipulation
• Reduces computational costs while preserving background structural consistency

Innovation

• Reversible Inversion two-stage denoising process
• Mask-Guided Selective Denoising for target regions
• Training-free method with low computational overhead