EditThinker: Unlocking Iterative Reasoning for Any Image Editor

TL;DR

EditThinker introduces an iterative, deliberative framework for instruction-based image editing, enabling models to critique and refine edits through multiple cycles, significantly enhancing instruction-following accuracy.

Contribution

The paper presents a novel deliberative editing framework with a single multi-modal large language model that iteratively critiques and refines image editing instructions, improving performance over existing methods.

Findings

Significant improvement in instruction-following accuracy across four benchmarks.

Effective use of reinforcement learning to align reasoning with editing actions.

Demonstrated generalization to any image editing model.

Abstract

Instruction-based image editing has emerged as a prominent research area, which, benefiting from image generation foundation models, have achieved high aesthetic quality, making instruction-following capability the primary challenge. Existing approaches improve instruction adherence via supervised or reinforcement learning, yet single-turn success rates remain limited due to inherent stochasticity and a lack of deliberation. In this work, we propose a deliberative editing framework to 'think' while they edit, which simulates the human cognitive loop by iteratively executing a Think-while-Edit cycle: Critiquing results and Refining instructions , followed by Repeating the generation until satisfactory. Specifically, we train a single MLLM, EditThinker, to act as the reasoning engine of this framework, which jointly produce the critique score, reasoning process, and refined instructions. We employ reinforcement learning to align the EditThinker's thinking with its editing, thereby generating more targeted instruction improvements. Extensive experiments on four benchmarks demonstrate that our approach significantly improves the instruction-following capability of any image editing model by a large margin. We will release our data construction framework, datasets, and models to benefit the community.