PyTorch implementation of our VCR-Net models [pdf].
Environment installation as follows Segformer
- python 3.9
pip install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html
pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu111/torch1.9.0/index.html
pip install scipy
pip install einops
pip install torchcontrib
pip install ipython
pip install attr
pip install termcolor
pip install transformers
pip install scikit-image
pip install scikit-learn- You can download the CAL dataset from [ Onedrive | Baidu Pan(q9bh) ]. Download the dataset and place it in the dataset/ folder
You can test the trained model by running test.py.
Download the pretrained models from [ Onedrive | Baidu Pan(7itq) ] and place it in the pretrained/ folder
Download the Bert model from [ Onedrive | Baidu Pan(p468) ] and place it in the bert-base-cased/ folder
Download the models from [ Onedrive | Baidu Pan(2sd3) ] and place it in the save_models/ folder
python test.py