Ground truth missing from DRIVE test set?

https://drive.grand-challenge.org/

The dataset description states: "For the test cases, two manual segmentations are available; one is used as gold standard, the other one can be used to compare computer generated segmentations with those of an independent human observer. "

However these manual segmentations seem to be missing completely from the Dropbox link provided. Is there a reason for this?

Yes, those annotations are the thing that your algorithm needs to predict. You can compare your predictions to the gold standard by submitting your predictions to the challenge.