The pytvision package consists of my datasets, models, and image transformations methods for computer vision projects. This package also containing the synthetic render methods. The backend transforms using opencv.
pytorch 0.4.1
git clone https://github.com/pytorch/vision.git
cd vision
python setup.py installFrom source:
python setup.py build_ext --inplace
python setup.py installdocker build -f "Dockerfile" -t pytv:latest .
docker run -ti --privileged --ipc=host --name pytv-dev -p 8080:8080 -p 6006:6006 -v $DATASETS:/.datasets pytv:latest /bin/bashdocker-compose up --build -d
docker-compose down
docker exec -it pytv-dev /bin/bashThe full documentation is available at doc
- cars169
- cub2011
- StanfordOnlineProducts
- fer+
- inception_v4
- nasnetalarge
- dexpression
- unet (unet11, unetresnet, ...)
- pspnet
- Color transformation
- Geometrical transformation
- Blur + Noise transformation