Timeline
Here we summarize key dates and the competition timeline.
| Date | Event |
|---|---|
| 01/06/2026 | Data Challenge: The data submission portal opens |
| 15/06/2026 | Algorithm Challenge – Familiarization: The prediction submission portal opens |
| 25/06/2026 | MUnitQuest Symposium at ISEK2026 (12:30 - 14:00, CET) |
| 15/09/2026 | Data Challenge: The data submission closes (23:59, CET) |
| 30/09/2026 | Algorithm Challenge – Familiarization: The prediction submission portal closes |
| 15/10/2026 | Double-blind data evaluation completed |
| 01/11/2026 | Algorithm Challenge – Showdown: Starts by releasing the MUnitQuest data collection, prediction submission opens |
| 31/12/2026 | Algorithm Challenge – Showdown: The prediction submission portal (Isometric and Dynamic tasks) closes (23:59, CET) |
| 15/01/2027 | The awards are announced and final leaderboards are published 🏆 |
Data Challenge
- Community members collect experimental and simulated HDsEMG data together with motor unit spike train labels
- Datasets need to be submitted through the data submission portal (see Submission and Registration)
- An expert reviewer panel evaluates each dataset contribution (see Data Challenge for evaluation criteria)
- Outstanding data contributions (top 5, rated by an expert reviewer panel) are invited to contribute to a special issue in the Journal of Electromyography and Kinesiology
- After the competition is completed, data is released on an open data repository (mandatory for datasets entering the MUnitQuest data collection, optional otherwise)
Algorithm Challenge
Familiarization Phase
- Runs parallel to the Data Challenge
- Competitors are provided with training data from the MUniverse benchmark collection (including motor unit spike train labels)
- Competitors develop and optimize their algorithms
- Participants can compete in both tasks (Isometric and Dynamic) or in only one task
- Competitors need to upload their predictions (motor unit spike trains) on Codabench (coming soon)
- An anonymous preliminary leaderboard is provided as feedback
- Results from baseline algorithms (together with tutorials – coming soon) are provided for reference
- Participation in the Familiarization phase is encouraged
Showdown Phase
- Finals to determine the leaderboards of the Isometric and Dynamic tasks
- Competitors are provided with the MUnitQuest data collection (hidden labels) consisting of anonymized recordings collected during the Data Challenge
- Note that the organizers may make small alterations to the test data (e.g., injected noise, cropping signals, etc.) to limit the advantage of teams also participating in the Data Challenge
- Submission of predicted motor unit spike trains via the Codabench plattform
- To be eligible for awards, competitors need to share their code (e.g., using GitHub) openly