A version of the COCO JSON format with segmentation masks encoded with run-length encoding.
Below, learn the structure of COCO Run-Length Encoding (RLE).
{
"info": {
"description": "Exported from Darwin",
"url": "n/a",
"version": "n/a",
"year": 2022,
"contributor": "n/a",
"date_created": "2022/07/03"
},
"licenses": [
{
"url": "n/a",
"id": 0,
"name": "placeholder license"
}
],
"images": [
{
"license": 0,
"file_name": "2020-02-09--20-26-30(GMT-0700)--18s.mkv-0005.png1920x1080.png",
"coco_url": "n/a",
"height": 1080,
"width": 1920,
"date_captured": "",
"flickr_url": "n/a",
"darwin_url": "https://darwin.v7labs.com/api/images/437470172/original",
"darwin_workview_url": "https://darwin.v7labs.com/workview?dataset=484903&image=86",
"id": 290,
"tag_ids": []
}
],
"annotations": [
{
"id": 8944,
"image_id": 290,
"category_id": 8,
"segmentation": [
[
1550.78,
748.65,
1548.38,
770.21,
1638.96,
779.7,
1640.35,
757.28,
1640.83,
726.62,
1557.96,
718.72
]
],
"area": 4641.607800000347,
"bbox": [
1548.38,
718.72,
92.44999999999982,
60.98000000000002
],
"iscrowd": 0,
"extra": {}
},
]
}
With Roboflow supervision, an open source Python package with utilities for completing computer vision tasks, you can merge and split detections in COCO Run-Length Encoding (RLE). Read our dedicated guides to learn how to merge and split COCO Run-Length Encoding (RLE) detections.
Below, see model architectures that require data in the COCO Run-Length Encoding (RLE) format when training a new model.
On each page below, you can find links to our guides that show how to plot predictions from the model, and complete other common tasks like detecting small objects with the model.