nm000195
Mixture of LLP and EM for a visual matrix speller (ERP) dataset from
EEG Zarr
EEG recordings from an online brain-computer interface study with twelve healthy participants using a visual matrix P300 speller. Three fully unsupervised decoding methods were compared: learning from label proportions (LLP), expectation-maximization (EM), and their mixture (MIX). Participants copy-spelled a German sentence across three blocks, one per algorithm. The dataset supports research on calibration-free ERP-based BCIs.
- Participants
- 12
- Channels
- 31 (10-10)
- Citations
- 27
- HED
- v8.4.0
- Size
- 4.81 GB
- Version
- v1.0.3
- Updated
- Oct 8, 2026