https://github.com/Rock-100/MonoDet
[ICCV21 & WACV23] Monocular 3D Object Detection for Automonous Driving
https://github.com/Rock-100/MonoDet
3d-object-detection
Last synced: about 1 year ago
JSON representation
[ICCV21 & WACV23] Monocular 3D Object Detection for Automonous Driving
- Host: GitHub
- URL: https://github.com/Rock-100/MonoDet
- Owner: Rock-100
- License: apache-2.0
- Created: 2021-04-10T01:52:10.000Z (about 5 years ago)
- Default Branch: main
- Last Pushed: 2023-07-06T09:44:05.000Z (almost 3 years ago)
- Last Synced: 2025-03-19T06:03:46.776Z (about 1 year ago)
- Topics: 3d-object-detection
- Language: Python
- Homepage:
- Size: 11 MB
- Stars: 127
- Watchers: 6
- Forks: 21
- Open Issues: 5
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# MonoRCNN
MonoRCNN is a monocular 3D object detection method for autonomous driving, published at ICCV 2021 and WACV 2023. This project is an implementation of MonoRCNN.
## Related Link
* [Video demo (KITTI)](https://www.youtube.com/watch?v=1G_yV1nPoWE)
* [Paper (ICCV 2021)](https://arxiv.org/abs/2104.03775)
* [Paper (WACV 2023)](https://openaccess.thecvf.com/content/WACV2023/html/Shi_Multivariate_Probabilistic_Monocular_3D_Object_Detection_WACV_2023_paper.html)
* [KITTI benchmark (ICCV 2021)](http://www.cvlibs.net/datasets/kitti/eval_object_detail.php?&result=06e0aacbeb8de57a09d2a3557907064aa501048f)
* [KITTI benchmark (WACV 2023)](https://www.cvlibs.net/datasets/kitti/eval_object_detail.php?&result=5bbb320aee42eee314697370b8183217d61e8141)
## Visualization
* KITTI
* WAYMO
## Installation
* Python 3.6
* PyTorch 1.5.0
* Detectron2 0.1.3
Please use the Detectron2 included in this project. To ignore fully occluded objects during training, [`build.py`](detectron2/data/build.py#L55), [`rpn.py`](detectron2/modeling/proposal_generator/rpn.py#L292), and [`roi_heads.py`](detectron2/modeling/roi_heads/roi_heads.py#L272) have been modified.
## Dataset Preparation
* [KITTI](projects/KITTI/README.md)
## Model & Log
* [KITTI val split](https://drive.google.com/file/d/1OEHSLqg7goXb483GRLvM-z96ZDzN7wdI/view?usp=sharing)
Organize the downloaded files as follows:
```
├── projects
│ ├── MonoRCNN
│ │ ├── output
│ │ │ ├── model
│ │ │ ├── log.txt
│ │ │ ├── ...
```
* [Waymo](https://drive.google.com/file/d/1fL3E3yqZbRd3va3kN-1j-1YlX61-Qo13/view?usp=sharing)
## Test
```
cd projects/MonoRCNN
./main.py --config-file config/MonoRCNN_KITTI.yaml --num-gpus 1 --resume --eval-only
```
Set [`VISUALIZE`](projects/MonoRCNN/config/MonoRCNN_KITTI.yaml#L13) as `True` to visualize 3D object detection results (saved in `output/evaluation/test/visualization`).
## Training
```
cd projects/MonoRCNN
./main.py --config-file config/MonoRCNN_KITTI.yaml --num-gpus 1
```
## Citation
If you find this project useful in your research, please cite:
```
@inproceedings{MonoRCNN_ICCV21,
title = {Geometry-based Distance Decomposition for Monocular 3D Object Detection},
author = {Xuepeng Shi and Qi Ye and
Xiaozhi Chen and Chuangrong Chen and
Zhixiang Chen and Tae-Kyun Kim},
booktitle = {ICCV},
year = {2021},
}
```
```
@inproceedings{MonoRCNN_WACV23,
title = {Multivariate Probabilistic Monocular 3D Object Detection},
author = {Xuepeng Shi and Zhixiang Chen and Tae-Kyun Kim},
booktitle = {WACV},
year = {2023},
}
```
## Acknowledgement
* [Detectron2](https://github.com/facebookresearch/detectron2)
* [M3D-RPN](https://github.com/garrickbrazil/M3D-RPN)
* [MMDetection](https://github.com/open-mmlab/mmdetection)