A follow-up challenge has started – visit JustRAIGS to participate.
The AIROGS challenge is again open for submissions indefinitely! The prizes have already been awarded, but new submissions will be shown on the leaderboards. To view the new maximum number of submissions per team, click the "Submit" button.You can read more about the outcomes of AIROGS in our IEEE TMI paper.
Hardware requirements¶
📅 6 January 2022
All algorithm submissions will have available an NVIDIA T4 GPU (16 GB) with 8 CPUs (32 GB).
One record on Zenodo¶
📅 24 December 2021
A new version of the train set on Zenodo has been published here. Before, there were two separate records for storage space limits, which have now been increased. The second record is now deprecated.
Additional training data update¶
📅 20 December 2021
We have just made a change in the rules of the challenge regarding the use of additional training data. We first stated that the use of additional training data is not allowed. We have now decided that only the use of additional fundus image data (including weights pretrained on fundus image data) is prohibited. Hence, pretraining on images such as those from ImageNet is allowed.
Preliminary test phase opened including a Docker container example¶
📅 1 December 2021
Submissions to the preliminary test phase are open. How to make a
submission is explained
here, which will only be possible if you have a
verified account. We only accept participation requests from verified
accounts. See the documentation on how to get verified.
Prize money¶
📅 1 December 2021
These will be the prizes for the AIROGS winners:
🥇1st place: 3000 euros.
🥈2nd place: 2000 euros.
🥉3rd place: 1000 euros.
Additional prize money may be made available.
Training data available¶
📅 1 December 2021
Today the training data has been released on Zenodo (part 1 and part 2). The labels are available in train_labels.csv in part 1 of the dataset. In the future, the dataset will also be available on the AWS Registry of Open Data.