README.md

---
license: cc-by-nc-sa-4.0
---

# NVIDIA Dynamic Scenes Dataset - Extrapolation Type

This dataset is proposed by [Novel View Synthesis](https://gorokee.github.io/jsyoon/dynamic_synth/). 
[NVFi](https://github.com/vLAR-group/NVFi) selected skating and truck scenes from it, and rearrange the scenes to enable future extrapolation evaluation, 
and it's used by [FreeGave](https://github.com/vLAR-group/FreeGave) and [TRACE](https://github.com/vLAR-group/TRACE).

## Structure
The structure of the dataset is as:
```
Dynamic Scenes
| - data
| | - Skating: data for Skating scene
| | | - train: serves as training data
| | | - val: used for evaluating novel view interpolation
| | | - test: used for evaluating future extrapolation
| | | - transforms_train.json: camera poses and other meta informations for training set
| | | - transforms_val.json: camera poses and other meta informations for novel view interpolation task
| | | - transforms_test.json: camera poses and other meta informations for future extrapolation task
| | | - points3d.ply: randomly initialized points for 3D Gaussians
| | - Truck: data for Truck scene
```


## Citation
If you find this dataset helpful, please consider cite:
```bibtex
@article{li2023nvfi,
  title={NVFi: Neural Velocity Fields for 3D Physics Learning from Dynamic Videos}, 
  author={Jinxi Li and Ziyang Song and Bo Yang},
  year={2023},
  journal={NeurIPS}
}
```

Also please cite the original data contributors:
```bibtex
@article{yoon2020dynamic,
title={Novel View Synthesis of Dynamic Scenes with Globally Coherent Depths from a Monocular Camera},
author={Yoon, Jae Shin and Kim, Kihwan and Gallo, Orazio and Park, Hyun Soo and Kautz, Jan},
booktitle={The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month={June},
year={2020}
}
}
```