Import labels when creating task

[edurenye]

My actions before raising this issue

• Read/searched the docs
• Searched past issues

Expected Behaviour

When creating a new task I expect to be able to upload a bunch of images with their labels in labelMe xml files like you can do using app.roboflow.ai.

Current Behaviour

It just imports the images.

Possible Solution

Steps to Reproduce (for bugs)

1. Have images with their labelMe xml files, one for each image.
2. Create a task.
3. Add the labels (would be nice to also get them from the xml files)
4. Upload all files from dataset folder containing images and xml files.
5. Submit task.

Context

I have a dataset partially labeled using labelMe I want to easily import it without having to rename id's and stuff to be able to import, and of course I don't want to import file by file, I want to import all at once.

Your Environment

• Git hash commit (git log -1): 07de714
• Docker version docker version (e.g. Docker 17.0.05): 19.03.11
• Are you using Docker Swarm or Kubernetes? No
• Operating System and version (e.g. Linux, Windows, MacOS): Ubuntu 20.04
• 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.

[fedex-coder]

I join to this enhancement request.
I have a .json COCO file with a bunch of labels and annotations in it.
I upload images, but when i try to upload annotations file, the upload fails untill i add labels manually.
It would be great if they could be added automatically once uploaded annotation file.

[getcontrol]

+1 this is a huge limitation of CVAT

[ysohma]

+1

[ActiveChooN]

Could be partly resolved by #3790 by importing the dataset as a project.