Datasets:
quanhaoli
commited on
Commit
·
be35937
1
Parent(s):
54eca5e
update readme
Browse files
README.md
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
---
|
| 4 |
+
[[Paper](https://arxiv.org/pdf/2503.16421)] [[Project Page](https://quanhaol.github.io/magicmotion-site/)] [[Github](https://github.com/quanhaol/MagicMotion)]
|
| 5 |
+
|
| 6 |
+
## 🧙 MagicData Dataset
|
| 7 |
+
### 1. Dataset Introduction
|
| 8 |
+
|
| 9 |
+
The MagicData Dataset, introduced in [MagicMotion](https://arxiv.org/pdf/2503.16421), is a comprehensive trajectory controllable video generation dataset with rich semantic annotations. This dataset features high-quality videos with precise segmentation masks and bounding boxes annotations, designed to advance the field of controllable video synthesis and understanding.
|
| 10 |
+
|
| 11 |
+
The dataset consists of 23K diverse videos, each meticulously annotated with both pixel-level segmentation masks and object-level bounding boxes. Each video in the dataset is carefully curated to ensure high visual quality and annotation accuracy, making it suitable for training state-of-the-art video generation and understanding models.
|
| 12 |
+
|
| 13 |
+
|
| 14 |
+
### 2. File Structure
|
| 15 |
+
```
|
| 16 |
+
MagicData
|
| 17 |
+
├── videos
|
| 18 |
+
│ ├── videoid_1.mp4
|
| 19 |
+
│ ├── videoid_2.mp4
|
| 20 |
+
│ ├── ...
|
| 21 |
+
├── masks
|
| 22 |
+
│ ├── videoid_1
|
| 23 |
+
│ │ ├── annotated_frame_00000.png
|
| 24 |
+
│ │ ├── annotated_frame_00001.png
|
| 25 |
+
│ │ ├── ...
|
| 26 |
+
│ ├── videoid_2
|
| 27 |
+
│ │ ├── ...
|
| 28 |
+
├── boxs
|
| 29 |
+
│ ├── videoid_1
|
| 30 |
+
│ │ ├── annotated_frame_00000.png
|
| 31 |
+
│ │ ├── annotated_frame_00001.png
|
| 32 |
+
│ │ ├── ...
|
| 33 |
+
│ ├── videoid_2
|
| 34 |
+
│ │ ├── ...
|
| 35 |
+
```
|
| 36 |
+
|
| 37 |
+
### 3. Useful scripts
|
| 38 |
+
- Data Extraction
|
| 39 |
+
```bash
|
| 40 |
+
sudo apt-get install git-lfs
|
| 41 |
+
git lfs install
|
| 42 |
+
git clone https://huggingface.co/datasets/quanhaol/MagicData
|
| 43 |
+
tar -xzvf boxs.tar.gz
|
| 44 |
+
cat masks.tar.gz.part* > masks.tar.gz
|
| 45 |
+
tar -xzvf masks.tar.gz
|
| 46 |
+
cat videos.zip.part_* > videos.zip
|
| 47 |
+
unzip videos.zip
|
| 48 |
+
```
|
| 49 |
+
|
| 50 |
+
## Citation
|
| 51 |
+
If you found this dataset useful, please cite our [paper](https://arxiv.org/pdf/2503.16421).
|
| 52 |
+
```bibtex
|
| 53 |
+
@article{li2025magicmotion,
|
| 54 |
+
title={Magicmotion: Controllable video generation with dense-to-sparse trajectory guidance},
|
| 55 |
+
author={Li, Quanhao and Xing, Zhen and Wang, Rui and Zhang, Hui and Dai, Qi and Wu, Zuxuan},
|
| 56 |
+
journal={arXiv preprint arXiv:2503.16421},
|
| 57 |
+
year={2025}
|
| 58 |
+
}
|
| 59 |
+
```
|
| 60 |
+
|
| 61 |
+
## Contact
|
| 62 |
+
|
| 63 |
+
[liqh24@m.fudan.edu.cn](liqh24@m.fudan.edu.cn)
|