README.md

metadata

pretty_name: FineBadmintonBenchmark
language:
  - en
  - zh
task_categories:
  - question-answering
annotations_creators:
  - expert-generated
multilinguality:
  - multilingual
license: apache-2.0
configs:
  - config_name: action_classification
    default: true
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_action_classification.json
  - config_name: action_localization
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_action_localization.json
  - config_name: action_prediction
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_action_prediction.json
  - config_name: hitcount
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_hitcount.json
  - config_name: hitting_comment
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_hitting_comment2_updated.json
  - config_name: hitting_localization
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_hitting_localization_new.json
  - config_name: hitting_prediction
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_hitting_prediction.json
  - config_name: moving_recognition
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_moving_recognization.json
  - config_name: pointer_recognition
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_pointer_recog_updated.json
  - config_name: round_comment
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_round_comment_updated.json
  - config_name: roundcount
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_roundcount.json
  - config_name: select_detail_count
    data_files:
      - split: test
        path: finebadmintonbenchmark/benchmark_select_detail_count_questions.json

FineBadmintonBenchmark

Fine-grained badminton video question answering benchmark.

Dataset Structure

• hf_video_clips_qa/: one clip per QA item, named by video_uid (for example video_000001.mp4).
• finebadmintonbenchmark/: annotation JSON files.
  • each item contains video_uid
  • each QA item maps to exactly one video clip through video_uid

Citation

@inproceedings{he2025finebadminton,
  title={Finebadminton: A multi-level dataset for fine-grained badminton video understanding},
  author={He, Xusheng and Liu, Wei and Ma, Shanshan and Liu, Qian and Ma, Chenghao and Wu, Jianlong},
  booktitle={Proceedings of the 33rd ACM International Conference on Multimedia},
  pages={12776--12783},
  year={2025}
}

@misc{he2025finebadminton_arxiv,
  title={Finebadminton: A multi-level dataset for fine-grained badminton video understanding},
  author={He, Xusheng and Liu, Wei and Ma, Shanshan and Liu, Qian and Ma, Chenghao and Wu, Jianlong},
  year={2025},
  eprint={2508.07554},
  archivePrefix={arXiv},
  primaryClass={cs.MM},
  url={https://arxiv.org/abs/2508.07554}
}