Prepare Datasets for Mask2Former

A dataset can be used by accessing DatasetCatalog for its data, or MetadataCatalog for its metadata (class names, etc). This document explains how to setup the builtin datasets so they can be used by the above APIs. Use Custom Datasets gives a deeper dive on how to use DatasetCatalog and MetadataCatalog, and how to add new datasets to them.

MaskFormer has builtin support for a few datasets. The datasets are assumed to exist in a directory specified by the environment variable DETECTRON2_DATASETS. Under this directory, detectron2 will look for datasets in the structure described below, if needed.

$DETECTRON2_DATASETS/
  ADEChallengeData2016/
  coco/
  cityscapes/
  mapillary_vistas/

You can set the location for builtin datasets by export DETECTRON2_DATASETS=/path/to/datasets. If left unset, the default is ./datasets relative to your current working directory.

The model zoo contains configs and models that use these builtin datasets.

Expected dataset structure for COCO:

coco/
  annotations/
    instances_{train,val}2017.json
    panoptic_{train,val}2017.json
  {train,val}2017/
    # image files that are mentioned in the corresponding json
  panoptic_{train,val}2017/  # png annotations
  panoptic_semseg_{train,val}2017/  # generated by the script mentioned below

Install panopticapi by:

pip install git+https://github.com/cocodataset/panopticapi.git

Then, run python datasets/prepare_coco_semantic_annos_from_panoptic_annos.py, to extract semantic annotations from panoptic annotations (only used for evaluation).

Expected dataset structure for cityscapes:

cityscapes/
  gtFine/
    train/
      aachen/
        color.png, instanceIds.png, labelIds.png, polygons.json,
        labelTrainIds.png
      ...
    val/
    test/
    # below are generated Cityscapes panoptic annotation
    cityscapes_panoptic_train.json
    cityscapes_panoptic_train/
    cityscapes_panoptic_val.json
    cityscapes_panoptic_val/
    cityscapes_panoptic_test.json
    cityscapes_panoptic_test/
  leftImg8bit/
    train/
    val/
    test/

Install cityscapes scripts by:

pip install git+https://github.com/mcordts/cityscapesScripts.git

Note: to create labelTrainIds.png, first prepare the above structure, then run cityscapesescript with:

CITYSCAPES_DATASET=/path/to/abovementioned/cityscapes python cityscapesscripts/preparation/createTrainIdLabelImgs.py

These files are not needed for instance segmentation.

Note: to generate Cityscapes panoptic dataset, run cityscapesescript with:

CITYSCAPES_DATASET=/path/to/abovementioned/cityscapes python cityscapesscripts/preparation/createPanopticImgs.py

These files are not needed for semantic and instance segmentation.

Expected dataset structure for ADE20k:

ADEChallengeData2016/
  images/
  annotations/
  objectInfo150.txt
  # download instance annotation
  annotations_instance/
  # generated by prepare_ade20k_sem_seg.py
  annotations_detectron2/
  # below are generated by prepare_ade20k_pan_seg.py
  ade20k_panoptic_{train,val}.json
  ade20k_panoptic_{train,val}/
  # below are generated by prepare_ade20k_ins_seg.py
  ade20k_instance_{train,val}.json

The directory annotations_detectron2 is generated by running python datasets/prepare_ade20k_sem_seg.py.

Install panopticapi by:

pip install git+https://github.com/cocodataset/panopticapi.git

Download the instance annotation from http://sceneparsing.csail.mit.edu/:

wget http://sceneparsing.csail.mit.edu/data/ChallengeData2017/annotations_instance.tar

Then, run python datasets/prepare_ade20k_pan_seg.py, to combine semantic and instance annotations for panoptic annotations.

And run python datasets/prepare_ade20k_ins_seg.py, to extract instance annotations in COCO format.

Expected dataset structure for Mapillary Vistas:

mapillary_vistas/
  training/
    images/
    instances/
    labels/
    panoptic/
  validation/
    images/
    instances/
    labels/
    panoptic/
  mapillary_vistas_instance_{train,val}.json  # generated by the script mentioned below

No preprocessing is needed for Mapillary Vistas on semantic and panoptic segmentation.

If you want to evaluate instance segmentation on Mapillary Vistas, run python datasets/prepare_mapillary_vistas_ins_seg.py to generate COCO-style instance annotations.

Expected dataset structure for YouTubeVIS 2019:

ytvis_2019/
  {train,valid,test}.json
  {train,valid,test}/
    Annotations/
    JPEGImages/

Expected dataset structure for YouTubeVIS 2021:

ytvis_2021/
  {train,valid,test}.json
  {train,valid,test}/
    Annotations/
    JPEGImages/