BrushEdit: All-In-One Image Inpainting and Editing

TL;DR

BrushEdit introduces a novel image editing framework that combines multimodal large language models and inpainting techniques to enable autonomous, interactive, and precise image modifications guided by free-form instructions.

Contribution

The paper presents a new inpainting-based instruction-guided editing paradigm that overcomes limitations of existing inversion and instruction-based methods by integrating MLLMs with inpainting models in an agent-cooperative system.

Findings

Effective in performing large modifications like object addition/removal.

Achieves superior performance on multiple metrics including mask preservation and editing coherence.

Enables user-friendly, interactive image editing with free-form instructions.

Abstract

Image editing has advanced significantly with the development of diffusion models using both inversion-based and instruction-based methods. However, current inversion-based approaches struggle with big modifications (e.g., adding or removing objects) due to the structured nature of inversion noise, which hinders substantial changes. Meanwhile, instruction-based methods often constrain users to black-box operations, limiting direct interaction for specifying editing regions and intensity. To address these limitations, we propose BrushEdit, a novel inpainting-based instruction-guided image editing paradigm, which leverages multimodal large language models (MLLMs) and image inpainting models to enable autonomous, user-friendly, and interactive free-form instruction editing. Specifically, we devise a system enabling free-form instruction editing by integrating MLLMs and a dual-branch image inpainting model in an agent-cooperative framework to perform editing category classification, main object identification, mask acquisition, and editing area inpainting. Extensive experiments show that our framework effectively combines MLLMs and inpainting models, achieving superior performance across seven metrics including mask region preservation and editing effect coherence.