-
Python >= 3.6
-
PyTorch >= 1.5
We can successfully reproduce the main results under the setting below:
Tesla A100 (40G): CUDA 11.0 + PyTorch 1.7.1 + torchvision 0.8.2
-
GCC >= 4.9
-
cython:
pip install -U cython
-
simplejson:
pip install simplejson
-
PyAV:
conda install av -c conda-forge
-
iopath:
pip install -U iopath
orconda install -c iopath iopath
-
psutil:
pip install psutil
-
OpenCV:
pip install opencv-python
-
pandas:
pip install pandas
-
tensorboard:
pip install tensorboard
-
moviepy: (optional, for visualizing video on tensorboard)
conda install -c conda-forge moviepy
orpip install moviepy
-
timm:
pip install timm==0.4.12
git clone https://github.com/facebookresearch/detectron2.git
python -m pip install -e detectron2
If have to reinstall, first:
rm -rf build/ **/*.so
After having the above dependencies, run:
git clone https://github.com/MCG-NJU/EVAD.git
cd EVAD
python setup.py build develop
Now the installation is finished.