README.md

metadata

configs:
  - config_name: default
    data_files:
      - split: eval
        path: data/eval-*
      - split: DrawBench_trimmed
        path: data/DrawBench_trimmed-*
      - split: DiffusionDB_trimmed
        path: data/DiffusionDB_trimmed-*
      - split: Realism
        path: data/Realism-*
      - split: ABC_trimmed
        path: data/ABC_trimmed-*
dataset_info:
  features:
    - name: prompt
      dtype: string
    - name: category
      dtype: string
    - name: source
      dtype: string
    - name: uid
      dtype: int32
  splits:
    - name: eval
      num_bytes: 24907
      num_examples: 197
    - name: DrawBench_trimmed
      num_bytes: 7673
      num_examples: 77
    - name: DiffusionDB_trimmed
      num_bytes: 8173
      num_examples: 40
    - name: Realism
      num_bytes: 5383
      num_examples: 40
    - name: ABC_trimmed
      num_bytes: 3678
      num_examples: 40
  download_size: 38022
  dataset_size: 49814

Dataset Card

Dataset in ImagenHub.

Citation

Please kindly cite our paper if you use our code, data, models or results:

@article{ku2023imagenhub,
  title={ImagenHub: Standardizing the evaluation of conditional image generation models},
  author={Max Ku and Tianle Li and Kai Zhang and Yujie Lu and Xingyu Fu and Wenwen Zhuang and Wenhu Chen},
  journal={arXiv preprint arXiv:2310.01596},
  year={2023}
}