3DGCQA: A Database for 3D AIGCs

Author: Yingjie Zhou / Email: zyj2000@sjtu.edu.cn / Institution: Shanghai Jiao Tong University

Github: zyj2000 / HuggingFace:zyj2000

Thank you for your interest in this project, please feel free to contact me by Email if you have any questions.

(This is a draft document that provides only a brief introduction to the contents of the 3DGCQA, and will be further refined later.)

0. File Structure

obj: folder, contains all AIGCs' mesh, UV map and MTL file.

3DGCQA-info.csv: csv file, contains the time consumed for each 3D AIGC generation. (Some of the durations are very long due to the generation algorithms getting stuck)

3DGCQA-mos.csv: csv file, contains subjective scores (alignment and quality) for each 3D AIGC.

1. How to Access the 3DGCQA Database

Baidu Yunpan (百度云盘)： is coming

HuggingFace: Link

More info is coming soon...

Citation

@article{zhou20243dgcqa,
  title={3DGCQA: A Quality Assessment Database for 3D AI-Generated Contents},
  author={Zhou, Yingjie and Zhang, Zicheng and Wen, Farong and Jia, Jun and Jiang, Yanwei and Liu, Xiaohong and Min, Xiongkuo and Zhai, Guangtao},
  journal={arXiv preprint arXiv:2409.07236},
  year={2024}
}