CreatiDesign Dataset

Overview

CreatiDesign Dataset is a large-scale, high-quality dataset for multi-conditional graphic design generation. It contains 400,000 synthetic graphic design samples, each richly annotated with multiple conditions—including primary visual elements (main subject images), secondary visual elements (decorative objects with spatial and semantic annotations), and textual elements (such as slogans or titles with layout information). This dataset was constructed using a fully automated pipeline.

Features

400K Diverse Samples: Spanning various real-world graphic design scenarios.
Multi-Conditional Annotations: Each sample includes primary images, semantic layouts, and text layouts.
Automatic Construction: Ensures scalability and high annotation quality.

Loading the Dataset

from datasets import load_dataset
dataset_repo = 'HuiZhang0812/CreatiDesign_dataset'
train_dataset = load_dataset(dataset_repo, split='train')

@article{zhang2025creatidesign,
  title={CreatiDesign: A Unified Multi-Conditional Diffusion Transformer for Creative Graphic Design},
  author={Zhang, Hui and Hong, Dexiang and Yang, Maoke and Chen, Yutao and Zhang, Zhao and Shao, Jie and Wu, Xinglong and Wu, Zuxuan and Jiang, Yu-Gang},
  journal={arXiv preprint arXiv:2505.19114},
  year={2025}
}

52 个文件

浏览文件

数据集版权信息

本数据集的许可证为 Apache License 2.0。如有违反相关条款，请联系 WEHUB，我们将及时处理。查看许可证