OpenBMI — Open-Source BCI Benchmarking

A standardized, reproducible benchmarking platform for brain-computer interfaces.

I led the technical development of OpenBMI, a widely used open-source platform for reproducible BCI research.

  • Standardized, reproducible, real-time pipelines for the MI, ERP, and SSVEP paradigms.
  • Designed the evaluation methodology and metrics that compare decoding models on equal footing.
  • Released the accompanying EEG dataset, published as a peer-reviewed data descriptor in GigaScience (~700 citations; 221 stars, 113 forks).