SMART-Editor: A Multi-Agent Framework for Human-Like Design Editing with Structural Integrity
Year: 2026
Type: conference-paper
Abstract: Despite significant progress in natural image editing with state-of-the-art MLLMs, compositional layout and content editing for structured visual domains (e.g., posters, websites) remains under-explored.We introduce SMART-Editor, a multi-agent framework for compositional editing for structured images like posters or websites.Unlike prior models that focus on isolated local edits, SMART-Editor maintains global coherence through two complementary strategies: Reward-Refine, an inferencetime reward-guided refinement method, and RewardDPO, a training-time preference optimization approach leveraging reward-aligned layout pairs.To evaluate performance, we introduce SMARTEdit-Bench, a benchmark of cascading multi-step edit instructions that are implicit in nature yet require layout and semanticconsistency preserving reasoning about edit order to preserve spatial and semantic consistency.Both automatic and human evaluations confirm that reward-guided planning produces semantically consistent and visually coherent edits, beyond what single-shot VLMs can generate.
Institution: University of Maryland, College Park
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7140140979