Discord

Checklist Cyprus AI Camp 2026 IOAI Selection Camp · Task 6

Multimodal Skin Lesion Classification

Classify dermoscopic skin-lesion images into 7 diagnoses using the image together with patient age, sex and lesion site.

  • Multimodal
  • Vision
  • Tabular
  • Multimodal classification (images + metadata)

The task

Contestants build a model that combines dermoscopic images with patient metadata to predict one of 7 diagnoses from the ISIC/HAM10000 dataset: melanoma, basal cell carcinoma, actinic keratosis, benign keratosis, melanocytic nevus, dermatofibroma and vascular lesion.

Data are given both as JPEG files and as preprocessed NumPy arrays.

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
4,000 training and 1,000 test 224×224 images (JPEG and .npy), train/test tabular arrays and CSVs (age, sex, anatomical site), class_info.json, sample_submission.csv.
You submit
CSV submission of predicted diagnoses (format in the competition data page).
Scoring
Weighted F1 over the 7 classes (points formula not stated on the competition page).
Rules
  • Pre-trained models and additional data are not allowed unless the task says otherwise.
  • No outside help; LLM chats, ChatGPT, Copilot and similar tools are prohibited (except Colab inline autocompletion).
  • Only the Kaggle website and Google Colab may be used (no local IDEs); online library documentation is allowed.
  • One Kaggle account per participant; leaderboard name '[City] Name Surname'.
  • Task-specific rules are in the organisers' baseline notebook (the baseline downloads ImageNet ResNet-18 weights).
Format
Cyprus AI Camp 2026 (IOAI 2026 selection, in person, 31 Jan–8 Feb 2026): Multimodal, released 4 Feb 2026, 09:00–12:00 EET, on Kaggle (two tasks per 3-hour session).

Details

Year
2026, Cyprus (in person)
Round
IOAI Selection Camp · Task 6
Language
English
License
Kaggle licence field varies by task (MIT; CC BY-NC-SA 4.0; 'Subject to Competition Rules'); see each task's notes, as stated by the source