SpatialVID-Long

<h1 align='center'>SpatialVID: A Large-Scale Video Dataset with Spatial Annotations</h1> <div align='center'> <a href='https://oiiiwjh.github.io/' target='_blank'>Jiahao Wang</a>^1* <a href='https://FelixYuan-YF.github.io/' target='_blank'>Yufeng Yuan</a>^1* <a href='https://github.com/zrj-cn' target='_blank'>Rujie Zheng</a>^1* <a href='https://linyou.github.io' target='_blank'>Youtian Lin</a>¹ <a href='https://ygaojiany.github.io' target='_blank'>Jian Gao</a>¹ <a href='https://linzhuo.xyz' target='_blank'>Lin-Zhuo Chen</a>¹ </div> <div align='center'> <a href='https://openreview.net/profile?id=~yajie_bao5' target='_blank'>Yajie Bao</a>¹ <a href='https://github.com/YeeZ93' target='_blank'>Yi Zhang</a>¹ <a href='https://github.com/ozchango' target='_blank'>Chang Zeng</a>¹ <a href='https://github.com/yxzhou217' target='_blank'>Yanxi Zhou</a>¹ <a href='https://www.xxlong.site/index.html' target='_blank'>Xiaoxiao Long</a>¹ <a href='http://zhuhao.cc/home/' target='_blank'>Hao Zhu</a>¹ </div> <div align='center'> <a href='http://zhaoxiangzhang.net/' target='_blank'>Zhaoxiang Zhang</a>² <a href='https://cite.nju.edu.cn/People/Faculty/20190621/i5054.html' target='_blank'>Xun Cao</a>¹ <a href='https://yoyo000.github.io/' target='_blank'>Yao Yao</a>^1† </div> <div align='center'> ¹Nanjing University ²Institute of Automation, Chinese Academy of Science </div> <div align='center'> *Equal Contribution †Corresponding Author </div> <div align="center"> <strong>CVPR 2026</strong> </div> <br> <div align="center"> <a href="https://nju-3dv.github.io/projects/SpatialVID/"><img src="https://img.shields.io/static/v1?label=SpatialVID&message=Project&color=purple"></a> <a href="https://arxiv.org/abs/2509.09676"><img src="https://img.shields.io/static/v1?label=Paper&message=Arxiv&color=red&logo=arxiv"></a> <a href="https://github.com/NJU-3DV/spatialVID"><img src="https://img.shields.io/static/v1?label=Code&message=Github&color=blue&logo=github"></a> <a href="https://huggingface.co/SpatialVID"><img src="https://img.shields.io/static/v1?label=Dataset&message=HuggingFace&color=yellow&logo=huggingface"></a> <a href="https://www.modelscope.cn/organization/SpatialVID"><img src="https://img.shields.io/static/v1?label=Dataset&message=ModelScope&color=4285F4&logo=data:image/webp;base64,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"></a> </div> # SpatialVID ## Directory Structure ``` DATA_PATH/ ├─ data/ │ ├─ train/ │ │ └─ SpatialVID_Long_metadata.csv ├─ videos/ │ ├─ group_0000/ │ │ ├─ 0a286a04-6a99-5168-a985-480c88ace17e.mp4 │ │ ├─ 0a51005d-b77a-56ef-8f01-3669cb5dc075.mp4 │ │ └─ ... │ ├─ group_0001/ │ └─ ... └─ README.md ``` ## Dataset Download You can download the entire SpatialVID dataset using the following command: ```bash hf download SpatialVID/SpatialVID-Long --repo-type dataset ``` The whole dataset is approximately **10.5TB** in size. We have split the dataset into **244 groups** for easier management. Each group contains approximately 40GB of video data, with naming conventions following the format `group_0***` (e.g., `group_0000`, `group_0001`). For downloading specific files (instead of the full dataset), please refer to the [`download_SpatialVID.py`](https://github.com/NJU-3DV/SpatialVID/blob/main/utils/download_SpatialVID.py) provided in our GitHub repository. ## Usage Guide ### 1. Unzipping Group Files After downloading the group files (in `.tar.gz` format), use the `tar` command to extract their contents. For example: ```bash # Extract video groups (if downloaded separately) tar -xzvf videos/group_0000.tar.gz tar -xzvf videos/group_0001.tar.gz tar -xzvf videos/group_0002.tar.gz ... ``` ### 2. Using the Metadata File The `SpatialVID_Long_metadata.csv` file contains comprehensive metadata for all video clips in the dataset. It can be loaded with `pandas` for easy filtering and analysis based on your research needs. #### Metadata Fields Explained | Field Name | Description | |---------------------|-----------------------------------------------------------------------------| | `id` | Unique identifier for each video clip (matches the annotation folder name). | | `group id` | Identifier of the group the video clip belongs to (e.g., `group_0000`). | | `video path` | Relative path to the video file (e.g., `videos/group_0001/xxx.mp4`). | | `num frames` | Total number of frames in the video clip. | | `fps` | Frames per second (FPS) of the video clip. | | `resolution` | Resolution of the video clip (e.g., `1920x1080`). | | `aesthetic score` | Subjective score evaluating the video’s visual aesthetics (0–10). | | `luminance score` | Score measuring the video’s overall brightness. | | `motion score` | Score quantifying the intensity of motion in the video (0–20). | | `ocr score` | Score indicating the proportion of text occupying the screen (0–1). | #### Example: Loading Metadata with Pandas ```python import pandas as pd # Load the metadata file metadata = pd.read_csv('data/train/SpatialVID_long_metadata.csv') # Example: Filter videos with FPS ≥ 24 filtered_metadata = metadata[metadata['fps'] >= 24] print(filtered_metadata.head()) ``` ## License SpatialVID is released under the **Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC-BY-NC-SA 4.0)**. By using this dataset, you agree to the following terms: 1. **Attribution**: You must credit the original source of the dataset. 2. **Non-Commercial Use**: The dataset may not be used for commercial purposes. 3. **ShareAlike**: Any modified or derived works must be released under the same license. For the full license text, visit: https://creativecommons.org/licenses/by-nc-sa/4.0/legalcode. If you are the copyright owner of any video in our dataset and you need it to be removed, please contact us, and we will remove the video samples from our dataset / Github / project webpage / technical presentation as soon as possible. ## Citation If you use SpatialVID in your research, please cite our work using the following format (citation will be updated upon paper publication): ```bibtex @article{wang2025spatialvid, title={Spatialvid: A large-scale video dataset with spatial annotations}, author={Wang, Jiahao and Yuan, Yufeng and Zheng, Rujie and Lin, Youtian and Gao, Jian and Chen, Lin-Zhuo and Bao, Yajie and Zhang, Yi and Zeng, Chang and Zhou, Yanxi and others}, journal={arXiv preprint arXiv:2509.09676}, year={2025} } ```