GeoAI Arctic Challenge

Resources

A central hub for help, learning materials, and policies for the GeoAI Arctic Challenge.

๐Ÿ’ก Tip: If youโ€™re new here, start with the Quick Links below.

Contact Us

Tutorials & Learning Materials

Official

  • Starter materials: The Codabench files include helper scripts for loading images, inspecting labels, validating submissions, and encoding masks.
  • Submission format mini-guide: See Participate -> Submission Format for the COCO results JSON schema.

External

  • PyTorch segmentation recipes (Instance/semantic fundamentals) link
  • Detectron2 / MaskRCNN quickstarts (for instance masks) link
  • COCO mask utilities (pycocotools API reference) link
  • NumPy array loading (.npz files) link

Terms & Conditions

  • Data Usage
    • The dataset is provided for research and competition purposes within this challenge.
    • Redistribution of the dataset or derivatives outside the challenge requires permission from the original data owners and curators.
  • Attribution
    • Works or publications using this dataset should cite:
      Yang et al., 2023 (source dataset) and Li et al., 2025 (related RTS modeling work) (see citations below).
  • Submissions
    • Predictions must follow the required submission format and reflect automated model outputs (no manual labeling of test sets).
    • External data usage, if used, must be documented in your method description.
  • Teams & Conduct
    • Each team may include 1 to 5 members.
    • Do not share labels or predictions across different teams.
    • Follow the Codabench code of conduct and site terms.
  • Privacy
    • Team name, institution, and scores may appear on the public leaderboard.
  • Disqualification
    • Violations of the rules, data misuse, or attempts to game the leaderboard may result in removal from the competition.

Questions about policy? Email wenwen@asu.edu, chiayuhsu@asu.edu

Useful External Resources

Frequently Asked Questions

Q1. Where do I download the data?

A: Register and receive approval on Codabench, then download the public release package from the Files section. You can also see the Dataset Page for data details.

Q2. Whatโ€™s the submission format?

A: COCO results-format JSON with compressed RLE masks. See Participate -> Submission Format.

Q3. Can I use external data?

A: Yes, but you must document it in your method description.