Checklist AICC 2025 Round 0 · Task 2
Find Brain Tumors
Classify brain scan images into four classes when only about 2% of the training images are labelled.
The task
Framed as diagnosing brain scans at the 'AICC Hospital', the task asks for a classifier that assigns each test image to one of four classes: no tumour or one of three tumour types.
Roughly 2% of the training images carry labels; the remaining training images are unlabelled. A pretrained ResNet18 may be used, and code for loading it is provided in the baseline.
Contestants predict the class of every test image.
Abridged by SOTA from the official materials. The official statement has the exact rules, and it wins wherever this summary differs.
At a glance
- You get
train.csv(image_id, label), train/train/ (training images, about 2% labelled) and test/test/ (test images).- You submit
- CSV with columns ID and prediction.
- Scoring
- Macro F1 score.
- Rules
- No manual labelling.
- No pretrained models other than ResNet18.
- The notebook, including training, must run in at most 20 minutes and fit within a P100 GPU (16 GB VRAM).
- Individual participation (maximum team size 1); at most 50 submissions per day.
- AICC contest rules (stated on each Kaggle rules page, not enforceable): no use of LLMs for writing code or getting task ideas; no internet use other than official library documentation and the contest platform; no communication with anyone during the contest; clarifications only via the #clarification-requests channel on the AICC Discord server.
- Format
- AICC Round 0, online on Kaggle, 4 Oct 2025 18:00 UTC – 5 Oct 2025 18:00 UTC.