Getting Started with the CellMap Segmentation Challenge

Welcome to the CellMap Segmentation Challenge! This guide will help you get started with understanding the challenge, exploring the data, and submitting your results.

Step 1: Understand the Challenge

The CellMap Segmentation Challenge invites participants to advance machine learning for electron microscopy (EM) segmentation. Using a diverse, meticulously annotated dataset of eFIB-SEM images, this challenge aims to improve the segmentation of organelles and subcellular structures.

To learn more:

• Visit the About Page for an overview of the challenge and its goals.
• Review the GitHub Repository for detailed technical information, guidelines, and examples.

Step 2: Explore and Access the Data

The datasets for this challenge feature nearly 300 annotated training crops derived from over 20 diverse eFIB-SEM datasets, representing more than 40 unique organelle classes.

To dive deeper:

• Check out the Collection DOI Writeup for a detailed description of the datasets, their origins, and annotation protocols.
• Learn about the Annotation Classes to understand the types of structures included in the datasets.

Participants can access the CellMap Segmentation Challenge data by following these steps:

1. Visit the GitHub Repository: Go to the official GitHub repository. This repository contains detailed instructions, scripts, and resources needed to download the data, train models, make predictions, and submit results.
2. Download the Data: Use the provided download scripts and follow the guidelines outlined in the repository documentation to retrieve the datasets.
3. Train Your Model: Refer to example code and tutorials available in the repository to train machine learning models on the provided data.
4. Make and Submit Predictions: Follow the prediction submission guidelines in the repository. Submissions will remain open until early 2026.
5. Participate and Stay Updated: Keep an eye on updates and announcements through the GitHub repository and related community channels. Subscribe for email updates here.

Step 3: Log in with Your GitHub Account

To participate in the challenge and make submissions, you must log in using your GitHub account:

1. Click "Log in" in the top right corner of any page.
2. Authenticate using your GitHub account.

Once logged in, you’ll be able to access submission features and manage your participation.

Step 4: Learn About Submissions

To participate, your submission must follow specific guidelines for file formats and evaluation criteria:

• Submission Data Format – Details on how to format your submission files.
• Evaluation Criteria – Understand how your submissions will be assessed.
• Evaluation Resampling – Learn how resampling is handled during evaluation.

Step 5: Make a Submission

When you're ready to submit:

1. Navigate to the Submission Page.
2. Log in with your GitHub account if you haven’t already.
3. Complete all required fields and upload your submission files in the correct format.
4. Confirm that your results comply with the Submission Data Format.

Step 6: Manage Your Submissions

Track, edit, or update your submissions via the My Submissions page. You can also manage team-related information here.

Step 7: Need Help?

If you have questions or run into technical issues, please reach out to our support team.

Our expectation is that the GitHub repository containing the segmentation model you developed and used to create the Predictions and the Data (including all relevant code, model weights, logs, validation metrics and anything else needed for someone else to access and use your model) will be a public repository under the 3-clause BSD license or another open source license. This is in service of our core belief in open science.