Noise Map Guidance: Inversion with Spatial Context for Real Image Editing

TL;DR

This paper introduces Noise Map Guidance (NMG), a new inversion method for real-image editing with diffusion models that preserves spatial context and editing quality without requiring optimization.

Contribution

NMG is a novel inversion approach that incorporates spatial context and eliminates the need for optimization, improving real-image editing with diffusion models.

Findings

NMG effectively preserves spatial context during inversion.

NMG is adaptable across various editing techniques.

NMG is robust to different DDIM inversion variants.

Abstract

Text-guided diffusion models have become a popular tool in image synthesis, known for producing high-quality and diverse images. However, their application to editing real images often encounters hurdles primarily due to the text condition deteriorating the reconstruction quality and subsequently affecting editing fidelity. Null-text Inversion (NTI) has made strides in this area, but it fails to capture spatial context and requires computationally intensive per-timestep optimization. Addressing these challenges, we present Noise Map Guidance (NMG), an inversion method rich in a spatial context, tailored for real-image editing. Significantly, NMG achieves this without necessitating optimization, yet preserves the editing quality. Our empirical investigations highlight NMG's adaptability across various editing techniques and its robustness to variants of DDIM inversions.