Please request the dataset by following the instruction here.

Participants will be ranked according to macro-averaged F1 score. However, for the completeness of scientific assessment, other metrics will be used, i.e. Unweighted Average Recall (UAR) and macro-AUC. In the case of ties, UAR will be used to rank the performance, then followed by macro-AUC. In the case where all metrics are tied, the time stamp of the submission will be used for the ranking, and the earlier submission will have higher ranking.

Participants will be ranked on these final metrics.

Submission file format (csv)
For each image in the test set, you must predict the probability of each class. An example of a submission file is shown below: 

image,none,infection,ischaemia,both
501000.jpg,0.130669757723808,0.783217191696167,0.042017094790936,0.044095922261477
501001.jpg,0.006754168774933,0.004397555720061,0.464470148086548,0.5243781208992
...
501500.jpg,0.083577610552311,0.886395275592804,0.006665781140327,0.023361394181848