Discover datasets

Preprocessed 64-channel EEG recordings from 8 healthy participants performing imagined speech of three vowel phonemes (a, i, u) in response to auditory and visual cues. The dataset was re-packaged in BIDS format from the MOABB collection and supports decoding of imagined speech with Riemannian manifold features and relevance vector machine classification.

Participants

8

Channels

60 (10-10)

Citations

97

HED

v8.4.0

Size

2.58 GB

Version

v1.0.3

Updated

Aug 18, 2026

EEG recordings from 24 healthy participants performing a multisensory (audio-visual) detection task. Participants detected targets embedded in simultaneous auditory and visual noise while selective attention varied. The dataset supports investigation of EEG spectral dynamics, including the spectral exponent, under changing attention and sensory statistics.

Participants

24

Channels

64 (10-10)

Citations

86

Size

23.1 GB

Version

v1.0.0

Updated

Jun 23, 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

EEG dataset from a passive auditory oddball (mismatch negativity) experiment conducted as part of the PURSUE project. Data were collected from participants at three primarily undergraduate institutions in Southern California, Massachusetts, and Virginia during 2017 and 2018. The task follows the ERP CORE paradigm.

Participants

281

Channels

32 (10-10)

Citations

1

Size

12.4 GB

Version

v1.0.0

Updated

Jun 3, 2026

ChineseEEG is a high-density EEG dataset with simultaneous eye-tracking recorded from 10 participants who silently read two Chinese children's novels (The Little Prince and Garnett Dream), with about 11 hours of reading per participant in total. It includes raw data, filtered and minimally pre-processed EEG derivatives, and BERT-base-chinese text embeddings of the stimulus materials. The resource supports research on semantic alignment between language model representations and neural activity, and on neural decoding.

Participants

10

Channels

128 (egi-geodesic)

Citations

23

Size

697 GB

Version

v1.0.0

Updated

Jun 26, 2026

A naturalistic multi-talker listening dataset in which 25 native Mandarin speakers underwent both 7T fMRI and 64-channel EEG while listening to the Chinese version of Le Petit Prince. Conditions included single-talker and mixed-talker speech with attention directed to different talkers, with comprehension quizzes after each run. The dataset supports research on speech perception, selective attention, and language processing under naturalistic conditions.

Participants

26

Channels

64

Size

405 GB

Version

v1.0.0

Updated

Jun 18, 2026

Scalp EEG recorded from 45 participants listening to a 12.4-minute audiobook recording of the first chapter of Alice's Adventures in Wonderland. The naturalistic paradigm was designed to study how the brain builds hierarchical syntactic structure and makes rapid, incremental linguistic predictions. The dataset includes word-level events annotated with lexical frequency, sound power, and n-gram, RNN, and context-free-grammar surprisal predictors, along with comprehension scores and the audio stimuli.

Participants

45

Channels

60 (other)

Citations

91

Size

3.88 GB

Version

v1.1.3

Updated

Jul 3, 2026

EEG dataset from a masked-priming-style ERP study examining neural associations between fingerspelling, printed English words, and American Sign Language signs in deaf readers. Participants viewed 300 prime-target pairs in which primes were English words, ASL signs, or fingerspelled words, followed by four-letter English word targets, while performing an animal-detection task. The data support investigation of cross-modal and cross-script priming effects in deaf readers.

Participants

24

Channels

32 (10-10)

Size

2.62 GB

Version

v1.0.0

Updated

Jun 27, 2026

MEG-MASC is a curated set of raw magnetoencephalography recordings from 27 English speakers who listened to about two hours of naturalistic stories drawn from the Manually Annotated Sub-Corpus (MASC). Each participant completed two sessions with four fictional stories, interleaved with random word lists and comprehension questions. Word and phoneme onsets and offsets are time-stamped in the event files, and the data are organized in BIDS. The dataset serves as a benchmark for large-scale encoding and decoding analyses of temporally resolved brain responses to speech.

Participants

27

Citations

52

Size

99.7 GB

Version

v1.0.1

Updated

Jun 3, 2026

EEG recordings from six healthy participants performing imagined speech of long words, comprising two word classes (cooperate and independent), with 64 channels recorded at 256 Hz. Trials were cued by auditory beeps and visual cues, and the data were analyzed with Riemannian manifold covariance features and relevance vector machines for brain-computer interface classification. The dataset is re-hosted in BIDS format from the MOABB benchmark collection.

Participants

6

Channels

60 (10-10)

Citations

98

HED

v8.4.0

Size

1.27 GB

Version

v1.0.3

Updated

Aug 18, 2026