Computer Vision News - July 2022

52 Medical Imaging AI tool FROM python:latest RUN pip3 install Keras RUN pip3 install numpy RUN pip3 install Pillow RUN pip3 install matplotlib RUN pip3 install tensorflow RUN mkdir ./Output ENV BASE _ DIR=. COPY infer.py . /infer.py COPY multi _ CNN.h5 ./multi _ CNN.h5 COPY Test ./Test We can now run the commands below to 1) build an image which contains our inference model, 2) run the container to produce and save the results. If everything has worked correctly, we should see something like the following message on our terminal window: docker build -t docker-dl-example -f Dockerfile . docker run docker-dl-example python3 infer.py After signing up to Docker Hub , it’s also possible to create a repository and upload an image. This is a registry where all available images are stored and can be pulled by other users. It’s also the place where you can search for images to write in your Dockerfile. Other basic commands to start with Docker: docker images -> to list the images on your local machine docker ps -a -> to list the containers running on your local machine docker rmi $ID -> to remove an image (provided one knows the ID) docker rm $ID -> to remove a container (provided one knows the ID) After your image is shared, other people can pull it and run it with the following commands: docker pull username/repository:tag #to run docker run username/repository:tag