IntrinsicEdit: Precise generative image manipulation in intrinsic space
This addresses the need for pixel-precise, versatile editing in generative AI for applications in digital content creation, though it builds incrementally on existing RGB-X diffusion frameworks.
The paper tackles the problem of precise control in generative image editing by introducing a workflow that operates in intrinsic-image latent space, achieving state-of-the-art performance across tasks like color/texture adjustments and object insertion/removal without additional data or fine-tuning.
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.