LabelBox hosts a free online "label it yourself" tool and is a popular solution for outsourced labeling. Their key differentiator is that you can bring your own labeling team. This is important for domains where labeling needs to be done by a specially trained expert (for example, labeling crop diseases or medical images).
You can export from LabelBox in two formats, CSV or JSON. No computer vision models will accept these formats directly. But fear not, Roboflow supports importing LabelBox JSON annotations and can convert them to any format your machine learning model requires.
Below, learn the structure of LabelBox JSON.
[{
"ID": "a9b7c5d3e1f",
"DataRow ID": "xy10z8a6b4c",
"Labeled Data": "https://storage.labelbox.com/IMG_001.JPG",
"Label": {
"helmet": [{
"geometry": [{
"x": 690,
"y": 1497
}, {
"x": 931,
"y": 1497
}, {
"x": 931,
"y": 1633
}, {
"x": 690,
"y": 1633
}]
}, {
"geometry": [{
"x": 1079,
"y": 1445
}, {
"x": 1324,
"y": 1445
}, {
"x": 1324,
"y": 1564
}, {
"x": 1079,
"y": 1564
}]
}]
},
"Project Name": "Boggle",
"External ID": "IMG_001.JPG"
}]With Roboflow supervision, an open source Python package with utilities for completing computer vision tasks, you can merge and split detections in LabelBox JSON. Read our dedicated guides to learn how to merge and split LabelBox JSON detections.
Below, see model architectures that require data in the LabelBox JSON 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.