ReCo-Data Dataset Card

Introduction

ReCo-Data is a large-scale, high-quality video editing dataset comprising 500K+ instruction-video pairs. This card provides its statistics, collection pipeline, and dataset format.

1. Dataset Statistics

Statistics

Figure Caption:

• (a) Overview of scale
• (b) Task distribution showing balanced quantities: Replace (156.6K), Style (130.6K), Remove (121.6K), and Add (115.6K). Human evaluation on 200 randomly sampled videos confirms that the proportion of high-quality data exceeds 90% for each task.
• (c) Details of video information

2. Collection Pipeline

Our data collection pipeline consists of six primary stages:

1. Raw data pre-processing: Filtering raw video data based on specific quality criteria.
2. Object segmentation: Extracting object masks from videos.
3. Instruction generation: Employing VLLM (i.e., Gemini-2.5-Flash-Thinking) to construct editing prompts.
4. Condition pair construction: Involving first frame editing and depth map generation to prepare the input conditions for VACE.
5. Video synthesis: Employing VACE to generate videos based on conditions.
6. Video filtering and re-captioning: Leveraging VLLM (i.e., Gemini-2.5-Flash-Thinking) again to filter out low-quality samples and re-caption the remaining videos.

The data synthesis process required approximately 76,800 GPU hours on NVIDIA RTX 4090, while the VLLM (i.e., Gemini-2.5-Flash-Thinking) operations incurred a total cost of approximately $13,600.

3. Dataset Format

Directory Structure

The dataset follows the directory structure below:

ReCo-Data/
├── add/
│   ├── add_data_configs.json
│   ├── src_videos/
│   │   ├── video1.mp4
│   │   └── ...
│   └── tar_videos/
│       ├── video1.mp4
│       └── ...
├── remove/
│   ├── remove_data_configs.json
│   ├── src_videos/
│   └── tar_videos/
├── replace/
│   ├── replace_data_configs.json
│   ├── src_videos/
│   └── tar_videos/
└── style/
    ├── style_data_configs.json
    ├── src_videos/
    │   ├── video1.mp4
    │   └── ...
    └── tar_videos/
        ├── video1-a_Van_Gogh_style.mp4
        └── ...

File Naming Conventions

• Add, Remove, Replace tasks: The file naming pattern is similar across these three tasks. Corresponding video pairs in src_videos and tar_videos folders share the same name.
• Style task: The naming format differs slightly. In the tar_videos folder, files use the -a_{}_style suffix to distinguish different style variations of the same source video.

4. Sample Usage

After downloading the dataset and the official repository, you can test and visualize samples using the following scripts:

Testing and Visualization (Single Task)

To test a specific task (e.g., the replace task):

python reco_data_test_single.py \
  --json_path ./ReCo-Data/replace/replace_data_configs.json \
  --video_folder ./ReCo-Data \
  --debug

Mixed Task Loading

You can also load a mixed dataset composed of the four tasks with arbitrary ratios:

python reco_data_test_mix_data.py \
  --json_folder ./ReCo-Data \
  --video_folder ./ReCo-Data \
  --debug

5. License

This dataset is licensed under the CC BY-NC-SA 4.0 (Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License).

The dataset is provided for academic research purposes only.

For any questions, please contact: zhwzhang@mail.ustc.edu.cn.

🌟 Star and Citation

If you find our work helpful for your research, please consider giving a star⭐ on this repository and citing our work.

@article{zhang2025region,
  title={Region-Constraint In-Context Generation for Instructional Video Editing},
  author={Zhang, Zhongwei and Long, Fuchen and Li, Wei and Qiu, Zhaofan and Liu, Wu and Yao, Ting and Mei, Tao},
  journal={arXiv preprint arXiv:2512.17650},
  year={2025}
}