Multi-agent collaborative pathways for Chinese traditional architectural image generation

A multi-agent collaborative framework for generating high-fidelity images of Chinese traditional architecture, such as those found along the Beijing Central Axis, significantly advances AI-generated content (AIGC) technology. This approach integrates specialized agents to interpret user intent, generate creative prompts, produce culturally accurate images, and assess aesthetic relevance. By leveraging a cultural knowledge base and collaborative workflows, the framework outperforms baseline models in handling vague inputs while ensuring cultural accuracy and creative diversity. The study demonstrates potential for digital preservation, creative transformation, and innovative cultural tourism experiences, offering new pathways for AI applications in heritage conservation.

Lu, Y., Yuan, W., Wang, M., Wang, P., Wu, S., Wu, J., Xing, W., Xie, W., & Yu, F. (2025). Multi-agent collaborative pathways for Chinese traditional architectural image generation. Scientific Reports, 15(1). https://doi.org/10.1038/s41598-025-18130-7