Awesome
Go Closer to See Better: Camouflaged Object Detection via Object Area Amplification and Figure-Ground Conversion
paper link: https://ieeexplore.ieee.org/document/10065514
prediction maps and pretrained model
- The prediction map of our method can be downloaded at https://drive.google.com/drive/folders/1rZ9IrbFz4dggik1vnK53PnNNy7l-1X_r?usp=sharing.
- The pretrained model of our model can be downloaded at https://drive.google.com/drive/folders/1n80O0RIAe4KT08SZG-UK6HsWgD_qouoQ?usp=sharing.
environment
- python = Python 3.8.13
- others packages can be found at requirement.txt
start
git clone https://github.com/Haozhe-Xing/SARNet.git
conda create --name myenv python=3.8.13
conda activate myenv
pip install -r requirements.txt
New: about the features map visualization!
if you want to visualize your features maps like the above imgages, you can user the code in folder "display_heatmpas".
if you want to visualize your model predictions like the above imgages, you can user the code in folder "display_heatmpas/combine.py".