Can we export the annotated dataset in Coco format ?
Can someone please confirm ?
Sorry, looks like I missed that question. I was going to look up the exact format that the Coco data is in, but it should be no problem. Prodigy will give you the pixel corrdinates (
(x, y)) of the boxes you draw on the image and the associated label, so all information you need is there and you just need to convert it.
If you need the center
(x, y) coordinates and the width and height of each box instead of the points, you can basically do this in reverse: https://prodi.gy/docs/computer-vision#tip-box-coordinates