Use Builtin Datasets

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.

Detectron2 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/
  coco/
  lvis/
  cityscapes/
  VOC20{07,12}/

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 instance/keypoint detection:

coco/
  annotations/
    instances_{train,val}2017.json
    person_keypoints_{train,val}2017.json
  {train,val}2017/
    # image files that are mentioned in the corresponding json

You can use the 2014 version of the dataset as well.

Some of the builtin tests (dev/run_*_tests.sh) uses a tiny version of the COCO dataset, which you can download with ./datasets/prepare_for_tests.sh.

Expected dataset structure for PanopticFPN:

Extract panoptic annotations from COCO website into the following structure:

coco/
  annotations/
    panoptic_{train,val}2017.json
  panoptic_{train,val}2017/  # png annotations
  panoptic_stuff_{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_panoptic_fpn.py, to extract semantic annotations from panoptic annotations.

Expected dataset structure for LVIS instance segmentation:

coco/
  {train,val,test}2017/
lvis/
  lvis_v0.5_{train,val}.json
  lvis_v0.5_image_info_test.json
  lvis_v1_{train,val}.json
  lvis_v1_image_info_test{,_challenge}.json

Install lvis-api by:

pip install git+https://github.com/lvis-dataset/lvis-api.git

To evaluate models trained on the COCO dataset using LVIS annotations, run python datasets/prepare_cocofied_lvis.py to prepare “cocofied” LVIS annotations.

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 Pascal VOC:

VOC20{07,12}/
  Annotations/
  ImageSets/
    Main/
      trainval.txt
      test.txt
      # train.txt or val.txt, if you use these splits
  JPEGImages/

Expected dataset structure for ADE20k Scene Parsing:

ADEChallengeData2016/
  annotations/
  annotations_detectron2/
  images/
  objectInfo150.txt

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