This dataset (WBCIC-SHU) contains multi-day EEG recordings from 51 healthy, right-handed, BCI-naive subjects performing a motor imagery brain-computer interface paradigm across three sessions per subject. Participants imagined left-hand, right-hand, or (for a subset of 11 subjects) foot movements in response to visual and auditory cues, yielding 39,600 trials in total. The dataset is intended for benchmarking motor imagery classification algorithms such as CSP, FBCSP, EEGNet, deepConvNet, and FBCNet, and was converted to BIDS format using MOABB.
- Participants
- 51
- Channels
- 59 (10-10)
- Citations
- 34
- HED
- v8.4.0
- Size
- 154 GB
- Version
- v1.0.5
- Updated
- Aug 18, 2026