The Mother of All BCI Benchmarks (MOABB) is a community-driven, open-source python toolkit designed to enable reproducible evaluation and comparison of brain–computer interface algorithms on a wide range of publicly available EEG datasets.
Key capabilities
- Build standardized benchmarks that evaluate popular BCI algorithms across many datasets and paradigms (ERP, SSVEP, cVEP, motor imagery, and more).
- Provide reference implementations and reproducible pipelines so results are comparable and auditable.
- Aggregate and publish algorithm rankings to give a clear, objective view of method performance.
- Facilitate faster, more transparent algorithm development and fair comparisons across the community.
Why use MOABB
- Open-source and actively maintained.
- Comprehensive dataset support and commonly used paradigms.
- Clear documentation and examples to reproduce experiments.
- Integrates with common ML and signal-processing libraries.
MOABB can easily be installed via pip:
pip install moabb
For usage examples, API reference, and tutorials, see the official documentation linked above.
The Data-Driven NeuroTechnology lab is actively supporting, maintaining, and contributing to the MOABB project.
