Create manifest file directly in S3.

[ztyree42]

My actions before raising this issue

• Read/searched the docs
• Searched past issues

Expected Behaviour

create.py should accept an s3 path in lieu of local file path.
create.py functionality should be accessible from UI.

Current Behaviour

User must download dataset locally -> create manifest -> upload to s3.

Possible Solution

Steps to Reproduce (for bugs)

Context

My organization has many datasets in s3 already so downloading each dataset just to be able to create the manifest.json is out of the question. s3fs is not a solution for files and folders already existing in s3 (it works fine if you're starting with an empty bucket and like that workflow). Moreover the AWS Registry of Open Data hints at a workflow wherein datasets never have to be downloaded locally but this seems out of reach with the current CVAT manifest schema. Finally, for those users attempting to host CVAT on EC2 and expose it to end users through the browser adding a new dataset to s3 now requires SSH access to the EC2 undermining the ability to allow users to create new annotation tasks (or add new datasets) without also giving SSH access, which I find extreme.

Your Environment

• Git hash commit (git log -1):
• Docker version docker version (e.g. Docker 17.0.05):
• Are you using Docker Swarm or Kubernetes?
• Operating System and version (e.g. Linux, Windows, MacOS):
• Code example or link to GitHub repo or gist to reproduce problem:
• Other diagnostic information / logs:
  Logs from `cvat` container

Next steps

You may join our Gitter channel for community support.