README.md

Setting Up Datasets

This directory contains symlinks to data locations.

Creating Symlinks for COCO

Symlink the COCO dataset:

ln -s /path/to/coco $DETECTRON/detectron/datasets/data/coco

We assume that your local COCO dataset copy at /path/to/coco has the following directory structure:

coco
|_ coco_train2014
|  |_ <im-1-name>.jpg
|  |_ ...
|  |_ <im-N-name>.jpg
|_ coco_val2014
|_ ...
|_ annotations
   |_ instances_train2014.json
   |_ ...

If that is not the case, you may need to do something similar to:

mkdir -p $DETECTRON/detectron/datasets/data/coco
ln -s /path/to/coco_train2014 $DETECTRON/detectron/datasets/data/coco/coco_train2014
ln -s /path/to/coco_val2014 $DETECTRON/detectron/datasets/data/coco/coco_val2014
ln -s /path/to/json/annotations $DETECTRON/detectron/datasets/data/coco/annotations

COCO Minival Annotations

Our custom minival and valminusminival annotations are available for download here. Please note that minival is exactly equivalent to the recently defined 2017 val set. Similarly, the union of valminusminival and the 2014 train is exactly equivalent to the 2017 train set. To complete installation of the COCO dataset, you will need to copy the minival and valminusminival json annotation files to the coco/annotations directory referenced above.

Creating Symlinks for PASCAL VOC

We assume that your symlinked detectron/datasets/data/VOC<year> directory has the following structure:

VOC<year>
|_ JPEGImages
|  |_ <im-1-name>.jpg
|  |_ ...
|  |_ <im-N-name>.jpg
|_ annotations
|  |_ voc_<year>_train.json
|  |_ voc_<year>_val.json
|  |_ ...
|_ VOCdevkit<year>

Create symlinks for VOC<year>:

mkdir -p $DETECTRON/detectron/datasets/data/VOC<year>
ln -s /path/to/VOC<year>/JPEGImages $DETECTRON/detectron/datasets/data/VOC<year>/JPEGImages
ln -s /path/to/VOC<year>/json/annotations $DETECTRON/detectron/datasets/data/VOC<year>/annotations
ln -s /path/to/VOC<year>/devkit $DETECTRON/detectron/datasets/data/VOC<year>/VOCdevkit<year>

PASCAL VOC Annotations in COCO Format

We expect PASCAL VOC annotations converted to COCO json format, which are available for download here.

Creating Symlinks for Cityscapes:

We assume that your symlinked detectron/datasets/data/cityscapes directory has the following structure:

cityscapes
|_ images
|  |_ <im-1-name>.jpg
|  |_ ...
|  |_ <im-N-name>.jpg
|_ annotations
|  |_ instanceonly_gtFile_train.json
|  |_ ...
|_ raw
   |_ gtFine
   |_ ...
   |_ README.md

Create symlinks for cityscapes:

mkdir -p $DETECTRON/detectron/datasets/data/cityscapes
ln -s /path/to/cityscapes/images $DETECTRON/detectron/datasets/data/cityscapes/images
ln -s /path/to/cityscapes/json/annotations $DETECTRON/detectron/datasets/data/cityscapes/annotations
ln -s /path/to/cityscapes/root $DETECTRON/detectron/datasets/data/cityscapes/raw

Cityscapes Annotations in COCO Format

We expect Cityscapes annotations converted to COCO json format, which we will make available for download soon.