Datasets:

Languages:
English
ArXiv:
License:
OCFBench / README.md
Gaaaavin's picture
Update README.md
ce14b22
---
license: cc-by-nc-sa-4.0
language:
- en
pretty_name: OCFBench
size_categories:
- 10K<n<100K
---
# Dataset Card for OCFBench
[[Paper]](https://arxiv.org/abs/2310.11239)
[[Code]](https://github.com/ai4ce/Occ4cast/)
[[Website]](https://ai4ce.github.io/Occ4cast/)
<!-- Provide a quick summary of the dataset. -->
The OCFBench dataset is curated in the paper [**Occ4cast: LiDAR-based 4D Occupancy Completion and Forecasting**](https://arxiv.org/abs/2310.11239).
The dataset is processed from public autonomous driving data to support the training and evaluation of the novel **occupancy completion and forecasting (OCF)** task.
# Uses
Please download each `.sqf` file from individual directories and mount them to local system for usage.
For larger files that are splited into several parts, please run the following code to merge the parts before mounting:
```
cat output_prefix_* > merged.sqf
```
Please refer to our [GitHub repository](https://github.com/ai4ce/Occ4cast/) for dataset structure and loading details.
## Citation
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
```
@article{Liu2023occ4cast,
title={LiDAR-based 4D Occupancy Completion and Forecasting},
author={Xinhao Liu and Moonjun Gong and Qi Fang and Haoyu Xie and Yiming Li and Hang Zhao and Chen Feng},
journal={arXiv preprint arXiv:2310.11239},
year={2023}
}
```