Discover datasets

A large EEG database from 87 healthy participants recorded during a single-day motor imagery brain-computer interface (BCI) experiment. The data are organized into three datasets (A, B, C) that share one Graz-protocol task, left- versus right-hand motor imagery, with 240 trials per participant (over 20,800 trials in total). It also provides demographic, personality and cognitive-trait information, which supports studies of how user profiles relate to BCI performance and cross-user classification algorithms.

Participants

87

Channels

27 (10-10)

Citations

44

HED

v8.4.0

Size

15.8 GB

Version

v1.0.5

Updated

Oct 6, 2026

A multichannel EEG dataset of self-paced (freewill) reach-and-grasp movements recorded from 23 healthy right-handed participants across 49 sessions. On each of 6808 trials the participant freely chose one of four cups, two filled with water and two empty, and decided when to move. Recordings were made at 250 Hz with a 31-channel scalp EEG montage plus EOG and auxiliary channels, to support brain-machine interface research on natural motor execution.

Participants

23

Channels

31 (10-10)

HED

v8.4.0

Size

2.53 GB

Version

v1.0.0

Updated

Oct 8, 2026

THINGS-EEG2 is a large EEG dataset from 10 participants who viewed 16,540 training images and 200 repeatedly presented test images from the THINGS database. Images were shown in rapid serial visual presentation at 5 Hz while 63-channel EEG was recorded at 1000 Hz across four sessions per subject. The dataset supports modeling of human visual object recognition and includes resting-state recordings, the authors' preprocessed epochs, and stimulus images.

Participants

10

Channels

63 (10-10)

Citations

88

Size

241 GB

Version

v1.1.0

Updated

Jul 5, 2026

EEG recordings from 13 healthy participants performing a copy-spelling task with a visual 6x7 matrix ERP speller. The speller was designed for learning from label proportions (LLP), an unsupervised, calibrationless classification approach for ERP-based brain-computer interfaces. Two stimulus sequences with different target/non-target ratios allowed LLP to reconstruct mean target and non-target responses without labeled data. In the original online study, 84.5% of characters were spelled correctly on average without prior calibration.

Participants

13

Channels

31 (10-10)

Citations

46

HED

v8.4.0

Size

5.15 GB

Version

v1.0.3

Updated

Oct 8, 2026

Stereo-EEG recordings from 10 Dutch-speaking patients with pharmaco-resistant epilepsy who read aloud 100 Dutch words presented one at a time on a laptop screen. Intracranial EEG (1103 recorded contacts spanning cortical and subcortical structures) was recorded simultaneously with the participants' pitch-shifted speech audio. The dataset supports research on speech production and speech decoding for brain-computer interfaces, and includes electrode localisations and FreeSurfer derivatives.

Participants

10

Size

9.20 GB

Version

v1.0.0

Updated

Oct 8, 2026

This dataset contains intracranial recordings acquired in humans using the CorTec BrainInterchange implantable device together with BCI2000 software. It supports the development of an ecosystem of technology and protocols for adaptive neuromodulation research. The data are shared in BIDS format with iEEG recordings.

Participants

3

Size

19.3 GB

Version

v1.0.0

Updated

Jun 25, 2026

EEG dataset investigating transcranial direct current stimulation (tDCS) modulation of visual cortex during motor imagery. Recordings were acquired with a Brain Products EEG system at 5000 Hz from a small group of healthy adult participants across three recording sessions. The dataset is organized in BIDS format.

Participants

5

Channels

3 (other)

Size

1.08 GB

Version

v1.0.0

Updated

May 13, 2026

3M-CPSEED is an EEG dataset recorded from 20 healthy native Chinese speakers performing Chinese Pinyin production in overt, silent-intended, and imagined speech conditions. The stimulus set of selected finals and initials covers key articulatory features of the Chinese phonological system. The dataset supports research on speech neurophysiology, neural differences across pinyin articulations, and transfer learning to other alphabetic languages.

Participants

20

Channels

32 (10-10)

Size

15.9 GB

Version

v1.0.0

Updated

Jun 29, 2026

This dataset contains EEG recordings collected while participants viewed visual and audiovisual stimuli in virtual reality depicting the Edzna Archaeological Site. It is intended to support research on how immersive, multisensory cultural-heritage environments shape brain activity. The data are organized in BIDS format with HED annotations.

Participants

102

Channels

8 (10-10)

Size

634 MB

Version

v1.0.0

Updated

May 13, 2026

Alljoined-1.6M is a large-scale EEG dataset of neural responses to rapid serial visual presentation of natural images from the THINGS database. It was recorded with a consumer-grade 32-channel EMOTIV FLEX2 system in 20 adult participants over four sessions each, yielding more than 1.6 million trials. The dataset was built to test whether deep-learning-based brain-computer interface and semantic decoding research can be done with affordable EEG hardware.

Participants

20

Channels

32 (10-10)

Size

8.78 GB

Version

v1.0.3

Updated

Sep 23, 2026