
Step-by-step installation:
# create a new conda environment.
conda create --name textfusenet python=3.7.3
conda activate textfusenet

# install pytorch 1.3.1.
conda install pytorch=1.3.1 torchvision cudatoolkit=10.1 -c pytorch

# install other libraries for building TextFuseNet-detectron2. You can get detailed versions from the requirements.txt, 
# and difference between different versions may lead to unknown influence on performance.
pip install opencv-python
pip install tensorboard
pip install yacs
pip install tqdm
pip install termcolor
pip install tabulate
pip install matplotlib
pip install cloudpickle
pip install wheel
pip install pycocotools

# download our code from gitHub.
git clone https://github.com/ying09/TextFuseNet.git
unzip TextFuseNet-master
cd TextFuseNet-master

# rebuild fvcore, we rebuild it for training more conveniently.
pip install fvcore-master.zip

# build   -detectron2, which needs cuda-10.1 and gcc-4.9.
python setup.py build develop

# If you have other problems about installation, please let us know on issue.
