ISIC 2016 challenge

Hi guys,

I am currently working on ISIC 2016 dataset currently. It would be great if anyone can respond to my query: The ground truth for the test set of ISIC 2016 dataset is annotated with 0 and 1 (for classification task 3) while the training set is annotated with benign and malignant labels. Could someone please tell if 0 corresponds to benign or malignant?

Thank you

For everyone’s reference, the ISIC 2016: Task 3 documentation states:

0.0 to 0.5: represent some confidence in the prediction that the lesion in the image in non-malignant (i.e. benign), with relatively lesser values indicating relatively more confidence in non-malignancy
0.5 to 1.0: represent some confidence in the prediction that the lesion in the image is malignant, with relatively greater values indicating relatively more confidence in malignancy

So, 0 is benign and 1 is malignant.