IntrinsicEdit: Precise generative image manipulation in intrinsic space

TL;DR

IntrinsicEdit introduces a versatile image editing method operating in an intrinsic-image latent space, enabling precise, local, and semantic manipulations across various tasks without additional data or fine-tuning.

Contribution

It presents a novel workflow that combines exact diffusion inversion and disentangled channel manipulation for high-precision image editing in intrinsic space.

Findings

Achieves state-of-the-art results in color and texture adjustments

Enables precise object insertion and removal

Handles complex relighting and combined edits effectively

Abstract

Generative diffusion models have advanced image editing with high-quality results and intuitive interfaces such as prompts and semantic drawing. However, these interfaces lack precise control, and the associated methods typically specialize on a single editing task. We introduce a versatile, generative workflow that operates in an intrinsic-image latent space, enabling semantic, local manipulation with pixel precision for a range of editing operations. Building atop the RGB-X diffusion framework, we address key challenges of identity preservation and intrinsic-channel entanglement. By incorporating exact diffusion inversion and disentangled channel manipulation, we enable precise, efficient editing with automatic resolution of global illumination effects -- all without additional data collection or model fine-tuning. We demonstrate state-of-the-art performance across a variety of tasks on complex images, including color and texture adjustments, object insertion and removal, global relighting, and their combinations.