85e5e798a8
|
||
|---|---|---|
| .. | ||
| car_assets | ||
| carassets | ||
| carla_dataset | ||
| data | ||
| models | ||
| textures | ||
| utils | ||
| weights | ||
| .gitattributes | ||
| .gitignore | ||
| 2.4.1' | ||
| LICENSE | ||
| data_loader.py | ||
| detect.py | ||
| generated_and_test.py | ||
| hubconf.py | ||
| nmr_test.py | ||
| readme.md | ||
| requirements.txt | ||
| test.py | ||
| train.py | ||
| train_camouflage_yolov3.py | ||
readme.md
The official implementation for AAAI2022 paper "FCA Learning a 3D Full-coverage Vehicle Camouflage for Multi view Physical Adversarial Attack"
Install the following python package before run the code.
neural_renderer
before you running the code, you must install the
neural rendererpython package. You can pull our implementation here, which is slight different to daniilidis.
If there is any question, please contact us without hesitate.
Note that, you code is based on Yolo-V3 implementation
After train the adversarial camouflage, you can see how camouflage like with the code in src folder.
carasset folder contains some necessary file.
carla_dataset folder contains the dataset, you can reference the carla_dataset/readme.md for detail.