I2E: From Image Pixels to Actionable Interactive Environments for Text-Guided Image Editing
This addresses limitations in existing image editing methods for tasks requiring precise local control and multi-object spatial reasoning, representing a novel paradigm shift rather than an incremental improvement.
The paper tackles the problem of compositional text-guided image editing by proposing I2E, a 'Decompose-then-Action' paradigm that transforms images into manipulable object layers and uses a physics-aware agent for atomic actions, resulting in significant outperformance over state-of-the-art methods on benchmarks like I2E-Bench in handling complex instructions and maintaining physical plausibility.
Existing text-guided image editing methods primarily rely on end-to-end pixel-level inpainting paradigm. Despite its success in simple scenarios, this paradigm still significantly struggles with compositional editing tasks that require precise local control and complex multi-object spatial reasoning. This paradigm is severely limited by 1) the implicit coupling of planning and execution, 2) the lack of object-level control granularity, and 3) the reliance on unstructured, pixel-centric modeling. To address these limitations, we propose I2E, a novel "Decompose-then-Action" paradigm that revisits image editing as an actionable interaction process within a structured environment. I2E utilizes a Decomposer to transform unstructured images into discrete, manipulable object layers and then introduces a physics-aware Vision-Language-Action Agent to parse complex instructions into a series of atomic actions via Chain-of-Thought reasoning. Further, we also construct I2E-Bench, a benchmark designed for multi-instance spatial reasoning and high-precision editing. Experimental results on I2E-Bench and multiple public benchmarks demonstrate that I2E significantly outperforms state-of-the-art methods in handling complex compositional instructions, maintaining physical plausibility, and ensuring multi-turn editing stability.