Marmot: Multi-Agent Reasoning for Multi-Object Self-Correcting in Improving Image-Text Alignment

While diffusion models excel at generating high-quality images, they often struggle with accurate counting, attributes, and spatial relationships in complex multi-object scenes. To address these challenges, we propose Marmot, a novel and generalizable framework that employs Multi-Agent Reasoning for Multi-Object Self-Correcting, enhancing image-text alignment and facilitating more coherent multi-object image editing. Our framework adopts a divide-and-conquer strategy that decomposes the self-correction task into three critical dimensions (counting, attributes, and spatial relationships), and further divided into object-level subtasks. We construct a multi-agent editing system featuring a decision-execution-verification mechanism, effectively mitigating inter-object interference and enhancing editing reliability. To resolve the problem of subtask integration, we propose a Pixel-Domain Stitching Smoother that employs mask-guided two-stage latent space optimization. This innovation enables parallel processing of subtask results, thereby enhancing runtime efficiency while eliminating multi-stage distortion accumulation. Extensive experiments demonstrate that Marmot significantly improves accuracy in object counting, attribute assignment, and spatial relationships for image generation tasks.

@article{sun2025_2504.20054,
  title={ Marmot: Multi-Agent Reasoning for Multi-Object Self-Correcting in Improving Image-Text Alignment },
  author={ Jiayang Sun and Hongbo Wang and Jie Cao and Huaibo Huang and Ran He },
  journal={arXiv preprint arXiv:2504.20054},
  year={ 2025 }
}