CVAT Setup

This project gives you a running instance of CVAT [Computer Vision Annotation Tool]. It automates all the steps needed for the setup.

Setup

• In order to be able to run this project, you need to install:

  • docker
  • docker-compose
  • python
  • git

• Create a file config.sh from config.sh.orig and fill in the variables found there:

  • CVAT_TARGET_HOST: the IP or host of the target machine; if the SSH port is not 22, you need to specify it like "127.0.0.1:2222"
  • CVAT_TARGET_USER: a user that can ssh into the target machine
  • CVAT_SSH_PRIVATE_KEY: the path to the private key

• Now run the setup.sh script:

  ./setup.sh
  

Usage

• Run the run.sh script

• You should now have a running instance of CVAT on the host you specified.

• Now create a superuser in CVAT. SSH into the remote machine and run:

  docker exec -it cvat bash -ic 'python3 ~/manage.py createsuperuser'