PromptSculptor: Multi-Agent Based Text-to-Image Prompt Optimization

2025-09-15

AI Summary

Manual prompt refinement in text-to-image generation is time-consuming and inefficient, requiring users to iteratively adjust prompts. Method: This paper proposes the first automated prompt optimization framework leveraging multi-agent collaboration and chain-of-thought (CoT) reasoning. It decomposes an ambiguous initial prompt into specialized subtasks—including semantic expansion, contextual completion, and style alignment—each executed by dedicated agents. A self-evaluation module and user-feedback-driven iterative refinement mechanism further enhance optimization. The framework is model-agnostic and seamlessly integrates with mainstream diffusion models (e.g., Stable Diffusion, SDXL, DALL·E). Contribution/Results: Experiments demonstrate a 62% reduction in average user iteration rounds compared to baselines, a 28% improvement in Fréchet Inception Distance (FID), and significantly higher user satisfaction. The approach exhibits strong generalizability and practical potential for industrial deployment.

Technology Category

• Natural Language Processing: Prompt Engineering / Prompting
• Search and Optimization: Learning to Search
• Computer Vision: Diffusion Models for Vision

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

The rapid advancement of generative AI has democratized access to powerful tools such as Text-to-Image models. However, to generate high-quality images, users must still craft detailed prompts specifying scene, style, and context-often through multiple rounds of refinement. We propose PromptSculptor, a novel multi-agent framework that automates this iterative prompt optimization process. Our system decomposes the task into four specialized agents that work collaboratively to transform a short, vague user prompt into a comprehensive, refined prompt. By leveraging Chain-of-Thought reasoning, our framework effectively infers hidden context and enriches scene and background details. To iteratively refine the prompt, a self-evaluation agent aligns the modified prompt with the original input, while a feedback-tuning agent incorporates user feedback for further refinement. Experimental results demonstrate that PromptSculptor significantly enhances output quality and reduces the number of iterations needed for user satisfaction. Moreover, its model-agnostic design allows seamless integration with various T2I models, paving the way for industrial applications.

Problem

• Automates iterative prompt optimization for text-to-image generation
• Transforms vague user prompts into comprehensive refined prompts
• Reduces iterations needed for high-quality image generation

Innovation

• Multi-agent framework automates iterative prompt optimization
• Chain-of-Thought reasoning infers hidden context details
• Model-agnostic design integrates with various T2I models