Diffusion-Based Conditional Image Editing through Optimized Inference with Guidance

TL;DR

This paper introduces a training-free, diffusion-based method for text-driven image editing that preserves source image structure while aligning with target prompts, using optimized inference with guidance.

Contribution

It proposes a novel guidance technique combining CLIP similarity and structural preservation to improve image translation without additional training.

Findings

Achieves high-quality image-to-image translation results.

Maintains source image structure effectively.

Works well across various tasks with pretrained models.

Abstract

We present a simple but effective training-free approach for text-driven image-to-image translation based on a pretrained text-to-image diffusion model. Our goal is to generate an image that aligns with the target task while preserving the structure and background of a source image. To this end, we derive the representation guidance with a combination of two objectives: maximizing the similarity to the target prompt based on the CLIP score and minimizing the structural distance to the source latent variable. This guidance improves the fidelity of the generated target image to the given target prompt while maintaining the structure integrity of the source image. To incorporate the representation guidance component, we optimize the target latent variable of diffusion model's reverse process with the guidance. Experimental results demonstrate that our method achieves outstanding image-to-image translation performance on various tasks when combined with the pretrained Stable Diffusion model.