Science Score: 44.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
✓CITATION.cff file
Found CITATION.cff file -
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
○DOI references
-
○Academic publication links
-
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (7.5%) to scientific vocabulary
Repository
Basic Info
- Host: GitHub
- Owner: yyl404
- License: apache-2.0
- Language: Python
- Default Branch: main
- Size: 13.3 MB
Statistics
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
- Releases: 0
Metadata Files
README.md
Install
conda create -n aal-det python=3.8
conda activate aal-det
pip install torch==2.1.0 torchvision==0.16.0 torchaudio==2.1.0 --index-url https://download.pytorch.org/whl/cu118
pip install mmcv==2.1.0 -f https://download.openmmlab.com/mmcv/dist/cu118/torch2.1/index.html
cd aal-det
pip install -e . -v
Dataset
For COCO dataset, you can just download official coco dataset and extract in data/coco
For Single-frame InfraRed Small Target(SIRST) dataset, run
shell
cd data
git clone https://github.com/YimianDai/open-sirst-v2.git
TRAIN
COCO
shell
python tools/train.py configs/aal/yolov8_coco_aal.py
python tools/test.py configs/aal/yolov8_coco_aal.py path/to/model.pth
SIRST
shell
python tools/train.py configs/aal/yolov8_sirst_aal.py
python tools/test.py configs/aal/yolov8_sirst_aal.py path/to/model.pth
SIRST
||mAP(clean)|mAP(FGSM)| |---|---|---| |YOLOv8-Clean|0.515|0.508| |YOLOv8-FGSM|0.535|0.489| |YOLOv8-AAL|0.496|0.499|
COCO
||mAP(clean)|mAP(FGSM)| |---|---|---| |YOLOv8-Clean||| |YOLOv8-FGSM||| |YOLOv8-AAL|||
Implementation
AAL method is supposed to improve trained models' test accuracy when the samples are perturbed by adversarial attack.
Available training and testing config files:
YOLOv8: AAL-Det/configs/aal/yolov8_coco_aal.py, AAL-Det/configs/aal/yolov8_sirst_aal.py
The AAL training loops are implemented in AAL-Det/mmdet/engine/runner/loops_adv.py.
NOTE: Although the complete AAL training requires backtracking process, we find that it unneccesary for detection models by experiments.
Owner
- Name: Yuanlong Yang
- Login: yyl404
- Kind: user
- Repositories: 1
- Profile: https://github.com/yyl404
Citation (CITATION.cff)
cff-version: 1.2.0 message: "If you use this software, please cite it as below." authors: - name: "MMDetection Contributors" title: "OpenMMLab Detection Toolbox and Benchmark" date-released: 2018-08-22 url: "https://github.com/open-mmlab/mmdetection" license: Apache-2.0
GitHub Events
Total
Last Year
Dependencies
- pytorch/pytorch ${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel build
- pytorch/pytorch ${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel build
- pytorch/pytorch ${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel build
- pytorch/pytorch ${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel build
- albumentations >=0.3.2
- cython *
- numpy *
- docutils ==0.16.0
- myst-parser *
- sphinx ==4.0.2
- sphinx-copybutton *
- sphinx_markdown_tables *
- sphinx_rtd_theme ==0.5.2
- urllib3 <2.0.0
- mmcv >=2.0.0rc4,<2.2.0
- mmengine >=0.7.1,<1.0.0
- fairscale *
- jsonlines *
- nltk *
- pycocoevalcap *
- transformers *
- cityscapesscripts *
- emoji *
- fairscale *
- imagecorruptions *
- scikit-learn *
- mmcv >=2.0.0rc4,<2.2.0
- mmengine >=0.7.1,<1.0.0
- scipy *
- torch *
- torchvision *
- urllib3 <2.0.0
- matplotlib *
- numpy *
- pycocotools *
- scipy *
- shapely *
- six *
- terminaltables *
- tqdm *
- asynctest * test
- cityscapesscripts * test
- codecov * test
- flake8 * test
- imagecorruptions * test
- instaboostfast * test
- interrogate * test
- isort ==4.3.21 test
- kwarray * test
- memory_profiler * test
- nltk * test
- onnx ==1.7.0 test
- onnxruntime >=1.8.0 test
- parameterized * test
- prettytable * test
- protobuf <=3.20.1 test
- psutil * test
- pytest * test
- transformers * test
- ubelt * test
- xdoctest >=0.10.0 test
- yapf * test
- mmpretrain *
- motmetrics *
- numpy <1.24.0
- scikit-learn *
- seaborn *