Computer Vision News - July 2022

53 Docker In our case, to check the obtained predictions, we need to first copy the output folder from the docker container into a local directory: docker ps -a -> to read the container ID for the image docker cp $CONTAINER ID:/Output ~/$local _ dir This article is a short introduction to Docker and shows the initial steps to set it up and use it for a medical imaging application. This tool shows great flexibility, from deploying web applications to deploying an app to sharing deep learning projects and let it run by many. And, as in the writer’s opinion, after spending countless hours setting up newmachines and conda environments, I would jump at any alternative to make this process less painful for developers and end users. Hence, I highly recommend looking into Docker. It might be your chance to reduce the risk of conflict between languages, libraries, or frameworks and run your models smoothly.