Discover datasets

EEG recordings from 20 healthy participants who wrote ten letters (a, d, e, f, j, n, o, s, t, v) with the right index finger after a visual cue. The dataset supports classification of handwritten characters, either directly from low-frequency and broadband EEG or with a two-step approach that first decodes continuous hand kinematics and then classifies the letter. It is intended for research on non-invasive brain-computer interfaces for communication.

Participants

20

Channels

60 (10-10)

Citations

13

HED

v8.4.0

Size

27.0 GB

Version

v1.0.2

Updated

Aug 18, 2026

The Finer-grained Affective Computing EEG Dataset (FACED) contains 32-channel scalp EEG from 123 healthy participants who watched 28 emotion-eliciting video clips targeting nine emotion categories (four negative, four positive, and neutral). Participants gave self-reported ratings on 12 dimensions, including discrete emotions, valence, arousal, liking, and familiarity. The dataset supports cross-subject affective computing and the development of EEG-based emotion recognition algorithms. This version is converted to BIDS with events annotated by video and emotion labels.

Participants

123

Channels

32 (10-10)

Citations

94

Size

31.4 GB

Version

v1.1.3

Updated

May 29, 2026

Beetl2021-A is a preprocessed four-class motor imagery EEG dataset (rest, left hand, right hand, feet) from Task 2 of the BEETL Competition at NeurIPS 2021. It contains 63-channel recordings at 500 Hz from healthy participants playing an online BCI racing game (Cybathlon2020IC), with 1490 trials of 4 s each in a single session. It serves as a benchmark for transfer learning and domain adaptation across heterogeneous EEG datasets and subjects, and is re-hosted in BIDS format via the MOABB pipeline.

Participants

3

Channels

63 (10-10)

HED

v8.4.0

Size

2.17 GB

Version

v1.0.6

Updated

Aug 20, 2026

EEG-BIDS dataset of cue-locked EEG recordings and event annotations from 43 neurotypical adults performing a random-dot kinematogram motion discrimination task. The task dissociates low-level priors (implicitly learned tone-motion contingencies) from high-level priors (a cover story about tinted glasses) against a neutral baseline. The data support analyses of hierarchical priors with signal detection theory, drift-diffusion modeling, and occipital oscillatory and aperiodic EEG activity, alongside psychosis-proneness and psychological flexibility questionnaires.

Participants

43

Channels

31 (10-10)

Size

8.76 GB

Version

v1.0.0

Updated

Jul 2, 2026

EEG dataset from a visual working memory experiment (Experiment 2) examining whether memory reactivation levels are affected by anticipated interference. Participants memorized 2 or 6 lateral objects, then performed either a dual task with distractors or a single fixation-only task during the delay, before selecting the cued object in a probe display. The data have been preprocessed; the preprocessing steps are described in the associated preprint and OSF repository (Experiment 2).

Participants

56

Channels

30 (10-10)

Size

40.1 GB

Version

v1.0.0

Updated

Jun 28, 2026

EEG recordings from participants viewing 200 object images presented in rapid serial visual presentation (RSVP) streams at 5 Hz or 20 Hz. The dataset supports investigation of the temporal representational dynamics of visual object processing under rapid presentation conditions. It accompanies a published NeuroImage study on object representations in RSVP streams.

Participants

16

Citations

97

Size

10.6 GB

Version

v1.0.0

Updated

May 13, 2026

This dataset contains 64-channel EEG recordings from 62 participants, organized into 31 pairs, who played a computerised version of the competitive Rock-Paper-Scissors game across 480 games. It was collected to study the neural correlates of competitive decision-making and to support decoding of choices and outcomes from brain activity. Each session lasted about one hour.

Participants

31

Channels

64

Citations

5

Size

78.0 GB

Version

v1.0.0

Updated

Jun 29, 2026

EEG dataset of multiband auditory brainstem responses (ABRs) derived from continuous 'peaky' speech, collected to identify optimal stimulus parameters. Experiment 1 compares zero-phase and chirp-phase speech profiles in normal-hearing adults, and Experiment 2 compares narrators with different fundamental frequencies (natural versus shifted to 100 Hz). Data include raw BrainVision EEG recordings, event files, and HDF5 stimulus and regressor files for deconvolution.

Participants

62

Channels

3 (other)

Size

94.3 GB

Version

v1.0.0

Updated

Jul 1, 2026

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

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