Solar Filament Segmentation Challenge 2026
Earth-Space AI Research
📌 About the Event
The Solar Filament Segmentation Challenge 2026 is an online computer-vision competition that challenges participants to develop algorithms for identifying and segmenting solar filaments in H-alpha images. Participants can use traditional image processing, machine learning, or deep-learning techniques. The challenge evaluates segmentation accuracy, structural consistency, computational efficiency, and code quality.
ℹ️ Event Details
- Type: Data Science and Computer Vision Competition
- Mode: Online
- Organizer/Host: Earth-Space AI Research (ESAIR) Lab
- Research Organizers: Researchers affiliated with the University of Missouri–St. Louis, Georgia State University, New Jersey Institute of Technology, and the National Solar Observatory
- Sponsor: U.S. National Science Foundation (NSF) and National Solar Observatory (NSO)
📅 Important Dates
- Competition Start: 10 July 2026
- Registration Deadline: 15 November 2026
- Winners Announcement: 30 November 2026
🎯 Eligibility & Rules
- Solutions may use traditional image processing, machine learning, or deep-learning approaches.
- Final evaluation requires a four-page technical report and a publicly accessible code repository.
- The repository must include the required dependencies and a notebook documenting the solution pipeline.
- Prize recipients must satisfy the organizers’ source-code, reproducibility, and payment requirements.
🏆 Benefits for Participants
- Compete for a share of the prize pool of up to US $3,000.
- Gain practical experience in computer vision, image segmentation, machine learning, and scientific data analysis.
- Build a portfolio project involving solar observations and space-weather research.
- Gain exposure to a research-oriented challenge associated with the IEEE Big Data Cup.