Discover datasets

License

License filtering isn't available while searching — clear the search box to filter by license.

Participants

Features

Has HED annotation
Has viewer
Has data quality Coming soon

The viewer and Zarr filters aren't available while searching — clear the search box to use them.

Channels & montage

Channel density Electrode system

Channel and montage filters apply while browsing — clear the search box to use them.

Reset

nm000246
Multi-day MI-BCI dataset (WBCIC-SHU) from Yang et al 2025

EEG Zarr

A multi-day motor imagery brain-computer interface EEG dataset recorded from 51 healthy, BCI-naive participants over three sessions on different days. Participants performed left-hand and right-hand motor imagery (2-class, 51 subjects), and a subset of 11 subjects also performed a 3-class paradigm that added foot-hooking. Recordings used 59 EEG channels plus ECG and EOG at 1000 Hz. The data are provided in BIDS format with HED event annotations and benchmark classification results (CSP, FBCSP, EEGNet, deepConvNet, FBCNet).

Participants

51

Channels

59 (10-10)

Citations

44

HED

v8.4.0

Size

58.4 GB

Version

v1.0.0

Updated

Jun 3, 2026

by Banghua Yang et al.
nm000262
P300 BCI EEG dataset (Chailloux Peguero et al. 2020)

EEG Zarr

A P300 brain-computer interface EEG dataset from 19 healthy participants, each recorded across 3 sessions. The study examines how visual stimulation conditions, standard flash versus cartoon face stimuli, affect single-option P300-BCI performance across grid sizes of 4 to 9 symbols. Recordings use an 8-channel montage sampled at 256 Hz and include target and non-target event labels.

Participants

19

Channels

8 (10-10)

Citations

16

HED

v8.4.0

Size

482 MB

Version

v1.0.2

Updated

Oct 8, 2026

by J. David Chailloux Peguero et al.
nm000342
Castillos et al. 2023, m-sequence c-VEP, 40% amplitude depth: Burst c-VEP Based BCI: Optimizing stimulus design for enhanced classification with minimal calibration data and improved user experience

EEG Zarr

EEG recordings from 12 healthy participants performing an offline 4-class code-modulated visual evoked potential (c-VEP) reactive BCI task. This subset contains the m-sequence condition at 40% amplitude depth, from a factorial study that compared burst and m-sequence codes at 100% and 40% modulation depth. The dataset supports evaluation of CNN-based decoding with minimal calibration data and of the visual comfort of reduced-intensity stimuli.

Participants

12

Channels

32 (10-10)

Citations

19

HED

v8.4.0

Size

340 MB

Version

v1.0.3

Updated

Oct 8, 2026

by Kalou Cabrera Castillos et al.
nm000345
Castillos et al. 2023, Burst c-VEP, 40% amplitude depth: Burst c-VEP Based BCI: Optimizing stimulus design for enhanced classification with minimal calibration data and improved user experience

EEG Zarr

EEG recordings from 12 healthy participants performing an offline 4-class reactive BCI task based on code-modulated visual evoked potentials (c-VEP). The dataset covers the 'Burst c-VEP' condition at 40% amplitude depth, in which short bursts of aperiodic flashes are presented at 2-4 flashes per second. It supports work on stimulus design, visual comfort, and CNN-based bitwise decoding with minimal calibration data.

Participants

12

Channels

32 (10-10)

Citations

18

HED

v8.4.0

Size

337 MB

Version

v1.0.3

Updated

Oct 8, 2026

by Kalou Cabrera Castillos et al.
on007720
BCI-FIT: A customization protocol for communication brain-computer interface systems.

EEG Zarr

EEG recordings from five participants with speech and/or physical impairments due to amyotrophic lateral sclerosis (ALS) who used a non-implantable communication brain-computer interface (cBCI) over multiple visits. Participants completed calibration and copy-spelling tasks with Matrix and RSVP layouts, in both customized and non-customized system configurations. Customization was iteratively adjusted based on typing performance, preferences, clinical observations, brain signal characteristics, and offline simulations. The data are organized in BIDS format.

Participants

5

Channels

7 (10-10)

Size

2.17 GB

Version

v1.0.0

Updated

Jun 19, 2026

by Betts Peters et al. OpenNeuro ds007720
nm000227
Eye-BCI Motor Execution dataset from Guttmann-Flury et al 2025

EEG Zarr

A BIDS-formatted EEG dataset of two-class hand-grasping motor tasks (left vs. right hand) from 31 healthy participants, recorded with a 64-channel Neuroscan system at 1000 Hz across up to three sessions of 40 trials each. It is a derivative of the multi-paradigm Eye-BCI dataset, which combines EEG, eye-tracking and high-speed video for ocular activity analysis. It is provided through MOABB for benchmarking motor-related brain-computer interface decoding.

Participants

31

Channels

66 (10-10)

Citations

13

HED

v8.4.0

Size

30.3 GB

Version

v1.0.2

Updated

Aug 18, 2026

by Eva Guttmann-Flury et al.
nm000184
BCI Competition 2020 Track 5 — ERP during walking (scalp, ear-EEG, and IMU)

EEG Zarr

EEG recordings from 15 healthy participants performing a visual P300 oddball task while walking on a treadmill at 1.6 m/s, released as BCI Competition 2020 Track 5. The dataset combines 46-channel scalp EEG, 14-channel ear-EEG, 4 EOG channels, and 6 inertial measurement unit channels sampled at 100 Hz. It is intended for evaluating brain-computer interface performance under ambulatory conditions, with trials divided temporally into training, validation, and test sets.

Participants

15

Channels

46 (10-10)

HED

v8.4.0

Size

224 MB

Version

v1.0.4

Updated

Oct 8, 2026

by Ji-Hoon Jeong et al.
nm000307
Perdikis2018: CNBI EPFL Cybathlon BCI-race motor-imagery dataset

EEG

EEG recordings from two tetraplegic pilots of team Brain Tweakers (CNBI, EPFL) during longitudinal motor-imagery BCI training for the Cybathlon 2016 BCI race. The dataset contains 16-channel EEG from cue-based offline calibration runs with three classes (both hands, both feet, rest), recorded at 512 Hz. It was re-released in BIDS format from the original Zenodo deposit via MOABB.

Participants

2

Channels

16 (10-10)

HED

v8.4.0

Size

5.30 GB

Version

v1.0.0

Updated

Oct 8, 2026

by Serafeim Perdikis et al.
nm000207
Kojima et al. 2024 (Dataset B) — Four-class ASME BCI: investigation of the feasibility and comparison of two strategies for multiclassing

EEG Zarr

EEG dataset of 15 healthy participants performing a four-class auditory stream segregation multiclass ERP (ASME) brain-computer interface task. Two strategies for achieving four classes were compared: ASME-4stream (four streams, one target stimulus each) and ASME-2stream (two streams, two target stimuli each). P300 responses, binary classification and offline BCI simulation with linear discriminant analysis showed average accuracies of 0.83 and 0.86, respectively. NASA-TLX workload ratings were also collected.

Participants

15

Channels

64 (10-10)

Citations

5

HED

v8.4.0

Size

23.5 GB

Version

v1.0.5

Updated

Oct 8, 2026

by Simon Kojima et al.
nm000313
Mainsah et al. 2025 — bigP3BCI: An Open, Diverse and Machine Learning Ready P300-based Brain-Computer Interface Dataset (Study S2)

EEG Zarr

This dataset contains EEG recordings from 24 healthy participants in Study S2 of bigP3BCI, a large open P300-based brain-computer interface collection. Participants used a 9x8 character grid speller with a house/tool stimulus paradigm, with target and non-target events annotated using HED. Data were recorded at 256 Hz from 32 channels and are provided in a machine-learning-ready BIDS format via MOABB.

Participants

24

Channels

32 (10-10)

Citations

2

HED

v8.4.0

Size

3.74 GB

Version

v1.0.3

Updated

Aug 18, 2026

by Boyla Mainsah et al.