Datasets:

Changli
/

DVD-Temp

license: apache-2.0
task_categories:
  - video-text-to-text
language:
  - en
pretty_name: DVD-Train
size_categories:
  - 10K<n<100K
tags:
  - video
  - caption
  - training
configs:
  - config_name: en
    data_files:
      - split: train
        path: data/en/caption.parquet

DVD-Train

Training set for Dialogue in ViDeos, containing detailed captions describing what happens in each video (including who said what).

Repository layout

DVD-Train/
├── data/
│   └── en/
│       └── caption.parquet    # English training captions
└── videos/
    └── en/
        └── *.mp4              # English training videos

Annotation schema

Each row in data/en/caption.parquet contains:

field	type	description
`video`	`string`	Video filename, e.g. `xxx.mp4`. The actual file is at `videos/en/<video>` in this repo.
`caption`	`string`	Detailed video description including scene, actions and dialogue-level speaker attribution.

Reference

If you find DVD-Train useful for your research, please cite our paper:

@article{tang2026dorca,
  title={{D-ORCA: Dialogue-Centric Optimization for Robust Audio-Visual Captioning}}, 
  author={Changli Tang and Tianyi Wang and Fengyun Rao and Jing LYU and Chao Zhang},
  journal={arXiv preprint arXiv:2602.07960},
  year={2026}
}