https://github.com/cuogeihong/ceasc
The official implementation of CEASC
https://github.com/cuogeihong/ceasc
Last synced: 5 months ago
JSON representation
The official implementation of CEASC
- Host: GitHub
- URL: https://github.com/cuogeihong/ceasc
- Owner: Cuogeihong
- License: apache-2.0
- Created: 2023-03-20T07:09:09.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2023-05-08T09:59:12.000Z (almost 2 years ago)
- Last Synced: 2024-10-28T04:19:30.936Z (6 months ago)
- Language: Python
- Size: 15 MB
- Stars: 105
- Watchers: 1
- Forks: 13
- Open Issues: 27
-
Metadata Files:
- Readme: README.md
- Contributing: .github/CONTRIBUTING.md
- License: LICENSE
- Code of conduct: .github/CODE_OF_CONDUCT.md
- Citation: CITATION.cff
Awesome Lists containing this project
- awesome-yolo-object-detection - CEASA - a4Wz04jLHFiAU88pUyDNQ)》" (Object Detection Applications)
README
# CEASC: Adaptive Sparse Convolutional Networks with Global Context Enhancement for Faster Object Detection on Drone Images
The repo is the official implementation of CEASC.
Our CEASC module is at [mmdet/models/dense_heads](mmdet/models/dense_heads)
Our Sparse Convolution Implementation is at [Sparse_conv](Sparse_conv)
Our config file is at [configs/UAV](configs/UAV)
## Requirement
Please follow [docs/en/get_started.md](docs/en/get_started.md) and install the [mmdetection](https://github.com/open-mmlab/mmdetection) toolbox.
a. Install [Pytorch 1.10.1](https://pytorch.org/)
b. Install [MMDetection](https://mmdetection.readthedocs.io/en/latest/) toolbox, required mmdet >= 2.7.0, mmcv-full >= 1.4.2.
- Our project utilizes mmdet == 2.24.1, mmcv-full == 1.5.1
c. Install albumentations and other packages.
```shell
pip install nltk
pip install -r requirements/albu.txt
```
d. Install our Sparse Convolution Implementation
```shell
cd ./Sparse_conv
python setup.py install
cd ..
```
## Usage
### 1. Data preparation
You could download VisDrone and UAVDT dataset (COCO Format) from official links or from other repositories like [UFPMP-Det](https://github.com/PuAnysh/UFPMP-Det).
### 2. Training
```shell
% training on a single GPU
python tools/train.py /path/to/config-file --work-dir /path/to/work-dir
% training on multi GPUs
bash tools/dist_train.sh /path/to/config-file num-gpus --work-dir /path/to/work-dir
```
#### Checkpoints:
We provide the following checkpoints:
- GFL v1 baseline, corresponding to [baseline_gfl_res18_visdrone](https://github.com/Cuogeihong/CEASC/blob/main/configs/UAV/baseline_gfl_res18_visdrone.py): [Google Drive](https://drive.google.com/drive/folders/1OaLSG0Y3bz1DH3R7HB4y-1LBRHpgQ55c?usp=sharing)
- GFL v1 CEASC, corresponding to [dynamic_gfl_res18_visdrone](https://github.com/Cuogeihong/CEASC/blob/main/configs/UAV/dynamic_gfl_res18_visdrone.py): [Google Drive](https://drive.google.com/drive/folders/1v7pby3LqmcIdDI52KKUQ43Ra3tBQdusR?usp=sharing)
- RetinaNet baseline, corresponding to [baseline_retinanet_res18_visdrone](https://github.com/Cuogeihong/CEASC/blob/main/configs/UAV/baseline_retinanet_res18_visdrone.py): [Google Drive](https://drive.google.com/drive/folders/1Ws5UQri07GGZo_PUyGjFBQA5cI3pjn2K?usp=sharing)
- RetinaNet CEASC, corresponding to [dynamic_retinanet_res18_visdrone](https://github.com/Cuogeihong/CEASC/blob/main/configs/UAV/dynamic_retinanet_res18_visdrone.py): [Google Drive](https://drive.google.com/drive/folders/1Gu0D5XULRkMEGNTGGKNj7X6-WiZs2a34?usp=sharing)
### 3. Test
```shell
python tools/test.py /path/to/config-file /path/to/work-dir/latest.pth --eval bbox
```
## Citation
If you find our paper or this project helps your research, please kindly consider citing our paper in your publication.
```
@misc{ceasc,
title={Adaptive Sparse Convolutional Networks with Global Context Enhancement for Faster Object Detection on Drone Images},
author={Bowei Du and Yecheng Huang and Jiaxin Chen and Di Huang},
year={2023},
eprint={2303.14488},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```