# Marmot: Object-Level Self-Correction via Multi-Agent Reasoning

arXiv: https://arxiv.org/abs/2504.20054 · HTML: https://mlsanity.com/paper/2504.20054/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 11 of its 54 references that are in the map.
- First posted on arXiv 10 Apr 2025.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)
- Diffusion models / [Image generation, text-to-image and editing](https://mlsanity.com/diffusion/image-generation-text-to-image-and-editing/)
- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 11 of 11)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [Auto-Encoding Variational Bayes](https://mlsanity.com/paper/1312.6114/) (2013)
- [Multimodal Chain-of-Thought Reasoning in Language Models](https://mlsanity.com/paper/2302.00923/) (2023)
- [MultiDiffusion: Fusing Diffusion Paths for Controlled Image Generation](https://mlsanity.com/paper/2302.08113/) (2023)
- [LLM-grounded Diffusion: Enhancing Prompt Understanding of Text-to-Image Diffusion Models with Large Language Models](https://mlsanity.com/paper/2305.13655/) (2023)
- [Inpaint Anything: Segment Anything Meets Image Inpainting](https://mlsanity.com/paper/2304.06790/) (2023)
- [Idea2Img: Iterative Self-Refinement with GPT-4V(ision) for Automatic Image Design and Generation](https://mlsanity.com/paper/2310.08541/) (2023)
- [DiffusionAgent: Navigating Expert Models for Agentic Image Generation](https://mlsanity.com/paper/2401.10061/) (2024)
- [Divide and Conquer: Language Models can Plan and Self-Correct for Compositional Text-to-Image Generation](https://mlsanity.com/paper/2401.15688/) (2024)

## Built on by (most cited first, 3 of 3)

- [Agentic Visual Generation: From Generative Models to Agentic Control](https://mlsanity.com/paper/2609.06758/) (2026)
- [Hallo4D: Multi-Modal Hallucination Mitigation for Consistent Spatio-Temporal Generation](https://mlsanity.com/paper/2607.12752/) (2026)
- [Multi-History-Step SDE Inversion for Image Editing with Superior Regional Awareness](https://mlsanity.com/paper/2609.06602/) (2026)

## Abstract (arXiv)

While diffusion models excel at generating high-quality images, they often struggle with accurate counting, attributes, and spatial relationships in complex multi-object scenes. One potential solution involves employing Multimodal Large Language Model (MLLM) as an AI agent to construct a self-correction framework. However, these approaches heavily rely on the capabilities of the MLLMs used, often fail to account for all objects within the image, and suffer from cumulative distortions during multi-round editing processes. To address these challenges, we propose Marmot, a novel and generalizable framework that leverages Multi-Agent Reasoning for Multi-Object Self-Correcting to enhance image-text alignment. First, we employ a large language model as an Object-Aware Agent to perform object-level divide-and-conquer, automatically decomposing self-correction tasks into object-centric subtasks based on image descriptions. For each subtask, we construct an Object Correction System featuring a decision-execution-verification mechanism that operates exclusively on a single object's segmentation mask or the bounding boxes of object pairs, effectively mitigating inter-object interference and enhancing editing reliability. To efficiently integrate correction results from subtasks while avoiding cumulative distortions from multi-stage editing, we propose a Pixel-Domain Stitching Smoother, which employs mask-guided two-stage latent space optimization. This innovation enables parallel processing of subtasks, significantly improving runtime efficiency while preventing distortion accumulation. Extensive experiments demonstrate that Marmot significantly improves accuracy in object counting, attribute assignment, and spatial relationships for image generation tasks.