EEGdash›NeMAR›NM000232

Iss. 232 · 10 subjects · 638 recordings · CC-BY 4.0

Dataset Brief · THINGS-EEG2

NM000232: eeg dataset, 10 subjects#

THINGS-EEG2: A large and rich EEG dataset for modeling human visual object recognition

Citation: Alessandro T. Gifford, Kshitij Dwivedi, Gemma Roig, Radoslaw M. Cichy (2022). THINGS-EEG2: A large and rich EEG dataset for modeling human visual object recognition. 10.17605/OSF.IO/3JK45

10-participant EEG dataset — THINGS-EEG2: A large and rich EEG dataset for modeling human visual object recognition.

Data & curation Alessandro T. Gifford · Kshitij Dwivedi · Gemma Roig · Radoslaw M. Cichy
Year 2022 · Distributed via NeMAR
Funding German Research Council (DFG) grants CI 241/1-1, CI 241/3-1, CI 241/1-7 · European Research Council (ERC) Starting Grant ERC-2018-StG 803370 · + 1 more

EEG · 63 ch1000 HzBIDS 1.9.05 tasks4 sessions

Layer 01Study

What was asked

Hypothesis, independent & dependent variables, paradigm, cohort, and the editorial caveats around what the recordings can and cannot answer.

Layer 02Signal · BIDS

What was recorded

Sidecars, channels & electrodes, coordinate system, event semantics, and quality stats from the NEMAR pipeline when available.

Layer 03Training · ML

What you can train on

Recommended access modes — MNE Raw, braindecode windows, PyTorch DataLoader — plus the targets the metadata makes addressable.

§ 01Access · Get started

Quickstart#

Get Started

Install

pip install eegdash

Access the data

from eegdash.dataset import NM000232

dataset = NM000232(cache_dir="./data")
# Get the raw object of the first recording
raw = dataset.datasets[0].raw
print(raw.info)

Query & Filter

Filter by subject

dataset = NM000232(cache_dir="./data", subject="01")

Advanced query

dataset = NM000232(
    cache_dir="./data",
    query={"subject": {"$in": ["01", "02"]}},
)

Iterate recordings

for rec in dataset:
    print(rec.subject, rec.raw.info['sfreq'])

Cite This Dataset

If you use this dataset in your research, please cite the original authors.

BibTeX

@dataset{nm000232,
  title = {THINGS-EEG2: A large and rich EEG dataset for modeling human visual object recognition},
  author = {Alessandro T. Gifford and Kshitij Dwivedi and Gemma Roig and Radoslaw M. Cichy},
  doi = {10.17605/OSF.IO/3JK45},
  url = {https://doi.org/10.17605/OSF.IO/3JK45},
}

§ 02Study · The README

About This Dataset#

THINGS-EEG2: A large and rich EEG dataset for modeling human visual object recognition

EEG dataset of 10 subjects who viewed 16,540 distinct training images and 200

test images (each repeated ~80 times) using rapid serial visual presentation (RSVP) at 5 Hz, recorded on a BrainVision actiCHamp system at 1000 Hz.

The source files store 63 EEG channels (the online reference electrode is not stored). Stimuli are drawn from the THINGS database (Hebart et al. 2019).

Each subject completed 4 separate sessions; each session contained:

5 training runs (~3,360 trials each) covering ~16,540 unique images

1 test run (~4,080 trials) of 200 images repeated 20× per session

2 resting-state runs (one before, one after the main experiment)

Total: ~32,540 training trials + ~16,000 test trials per subject across 4 sessions.

View full README

The source files store 63 EEG channels (the online reference electrode is not stored). Stimuli are drawn from the THINGS database (Hebart et al. 2019).

Each subject completed 4 separate sessions; each session contained:

5 training runs (~3,360 trials each) covering ~16,540 unique images

1 test run (~4,080 trials) of 200 images repeated 20× per session

2 resting-state runs (one before, one after the main experiment)

Total: ~32,540 training trials + ~16,000 test trials per subject across 4 sessions.

Recording setup

Manufacturer: Brain Products (actiCHamp)

63 EEG channels (one electrode served as online reference and is not stored in the source files)

10-10 cap layout

Sampling rate: 1000 Hz

Online band-pass: 0.01-100 Hz

Triggers recorded as BrainVision stimulus annotations (not as a dedicated stim channel)

Tasks (BIDS labels)

task-train: training run (RSVP of unique images)

task-test: test run (RSVP of repeated test images)

task-rest: resting state (eyes open, fixation cross)

Run numbering

task-train: run-01..run-05 per session (5 training parts)

task-test: single run per session

task-rest: run-01 (before main task) and run-02 (after main task)

Events

events.tsv columns:
onset, duration, sample, value, trial_type tot_img_number - global image ID (1-16540 for train; 1-200 for test;

‘n/a’ for target catch trials)

img_category - integer category index category_name - human-readable category, e.g. “01175_roller_coaster” block, sequence - hierarchical position within the run img_in_sequence - image position within its 20-image sequence soa - actual stimulus onset asynchrony (~200 ms)

trial_type values:
image - normal training/test image presentation target - random catch trial (subject must press a button) rest_marker - resting-state start/end marker

Subject information

participants.tsv contains age and sex (both extracted from the behavioural .mat files in the source data).

Folder layout

/sub-XX/ses-YY/eeg/ - main BIDS data (BDF + sidecars) /sourcedata/ - original BrainVision .eeg/.vhdr/.vmrk and

behavioural .mat files

/derivatives/preprocessed_eeg/ - authors’ preprocessed train/test epochs /derivatives/resting_state/ - authors’ preprocessed resting state /stimuli/ - image set (training_images.zip, test_images.zip)

plus image_metadata.npy

/code/ - this conversion script

Reference

Gifford, A.T., Dwivedi, K., Roig, G., & Cichy, R.M. (2022). A large and rich EEG dataset for modeling human visual object recognition. NeuroImage, 264, 119754. https://doi.org/10.1016/j.neuroimage.2022.119754

Code: gifale95/eeg_encoding OSF: https://osf.io/3jk45/

§ 03Cohort · Participants

Cohort#

Dataset Statistics#

Age distribution by gender (n=10, range 24–34 yr, mean 28.5 yr)

202530

Female · 8Male · 2

Sex composition

Female
8
Male
2
F : M ratio
4.00 : 1

80% female · n = 10 subjects with reported sex.

Channel counts: 63 ch (n=319 recordings)

Sampling frequencies: 1000.0 Hz (n=319 recordings)

Total recording duration: 87 h

§ 04Signal · Electrodes & trace

Signal · Electrodes & live trace#

Fig. 01 Signal & montage 63 ch · EEG · 1000 Hz · 10 subjects, 638 recordings

Live trace viewer — sub-08 · ses-02 · task-train

Showing one representative recording out of 10 subjects and 638 recordings in this dataset. Browse the full set on OpenNeuro; drop any other _eeg.{set,edf,bdf,vhdr} file onto the viewer (or pass ?eeg=<url>) to inspect it.

Electrode layout — EEG · 63 sensors — 63 channels

NEMAR Processing Statistics#

The plots below are generated by NEMAR’s automated EEG pipeline. The histogram shows pipeline success for data cleaning and ICA decomposition, the percentage of data frames and EEG channels retained after artefact removal, line noise per channel (RMS, dB), and the age/gender distribution of participants.

HED event descriptors word cloud

§ 05Manifest · BIDS tree

Manifest#

File Explorer#

Browse the BIDS file structure of this dataset. Records are fetched on demand from the EEGDash catalog the first time you open the explorer.

Recordings—

Files—

Subjects—

Modalities—

Click to load file structure…

§ 06API · Programmatic access

API Reference#

Signature

eegdash.dataset

class

eegdash.dataset.NM000232(cache_dir, query=None, s3_bucket=None, **kwargs)

Bases: EEGDashDataset

Author (year)Gifford2019

Canonical—

Importable asNM000232 · Gifford2019

Sourceeegdash/dataset/registry.py · [source ↗]

class eegdash.dataset.NM000232(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

THINGS-EEG2: A large and rich EEG dataset for modeling human visual object recognition

Study:: nm000232 (NeMAR)
Author (year):: Gifford2019
Canonical:: —

Also importable as: NM000232, Gifford2019.

Modality: eeg. Subjects: 10; recordings: 638; tasks: 5.

Parameters:

cache_dir (str | Path) – Directory where data are cached locally.
query (dict | None) – Additional MongoDB-style filters to AND with the dataset selection. Must not contain the key dataset.
s3_bucket (str | None) – Base S3 bucket used to locate the data.
**kwargs (dict) – Additional keyword arguments forwarded to EEGDashDataset.

data_dir#

Local dataset cache directory (cache_dir / dataset_id).

Type:: Path

query#

Merged query with the dataset filter applied.

Type:: dict

records#

Metadata records used to build the dataset, if pre-fetched.

Type:: list[dict] | None

Notes

Each item is a recording; recording-level metadata are available via dataset.description. query supports MongoDB-style filters on fields in ALLOWED_QUERY_FIELDS and is combined with the dataset filter. Dataset-specific caveats are not provided in the summary metadata.

References

OpenNeuro dataset: https://openneuro.org/datasets/nm000232 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000232 DOI: https://doi.org/10.17605/OSF.IO/3JK45

Examples

>>> from eegdash.dataset import NM000232
>>> dataset = NM000232(cache_dir="./data")
>>> recording = dataset[0]
>>> raw = recording.load()

__init__(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

save(path: str, overwrite: bool = False, offset: int = 0)[source]#

Save datasets to files by creating one subdirectory for each dataset:

path/
    0/
        0-raw.fif | 0-epo.fif
        description.json
        raw_preproc_kwargs.json (if raws were preprocessed)
        window_kwargs.json (if this is a windowed dataset)
        window_preproc_kwargs.json  (if windows were preprocessed)
        target_name.json (if target_name is not None and dataset is raw)
    1/
        1-raw.fif | 1-epo.fif
        description.json
        raw_preproc_kwargs.json (if raws were preprocessed)
        window_kwargs.json (if this is a windowed dataset)
        window_preproc_kwargs.json  (if windows were preprocessed)
        target_name.json (if target_name is not None and dataset is raw)

Parameters:

path (str) –

Directory in which subdirectories are created to store
-raw.fif | -epo.fif and .json files to.
overwrite (bool) – Whether to delete old subdirectories that will be saved to in this call.
offset (int) – If provided, the integer is added to the id of the dataset in the concat. This is useful in the setting of very large datasets, where one dataset has to be processed and saved at a time to account for its original position.

Access modesMNE → braindecode → PyTorch → ML

.rawMNE Raw object — standard tools (filter, epoch, ICA, plot_psd).mne

BaseConcatDatasetEach record is a lazy BaseDataset from braindecode — windowed via create_windows_from_events.braindecode

DataLoaderWraps the windowed dataset into a PyTorch DataLoader; supports parallel workers and on-the-fly augmentations.pytorch

Zarr cacheOptional braindecode Zarr mirror for fast resume; persisted to cache_dir.zarr

Hugging FaceNo per-dataset mirror published yet — browse the EEGDash org listing for sibling datasets. See the datasets loader API.huggingface

Croissant 1.0Machine-readable JSON-LD descriptor — NM000232.croissant.json (MLCommons schema, ingestible by PyTorch / TensorFlow / JAX).mlcommons

Examples using EEGDashcurated · start here

Find datasets with the EEGDash APIQuery the catalogue, filter by task or modality, list candidates.

Load one EEG recordingResolve a single record to an MNE Raw with channels and events.

EEG recording to PyTorch DataLoaderWrap braindecode windows in a DataLoader for model training.

Preprocess EEG and create windowsFilter, resample, epoch — and persist the windowed dataset.

Save and reload prepared dataCache a windowed dataset to disk and reattach it without recompute.

Download a dataset locallyPrefetch BIDS files to a local cache and validate the layout.

Swap any load_dataset(...) call for nm000232 to reproduce the tutorial on this dataset.

Citation

Alessandro T. Gifford, Kshitij Dwivedi, Gemma Roig, Radoslaw M. Cichy (2022). THINGS-EEG2: A large and rich EEG dataset for modeling human visual object recognition. 10.17605/OSF.IO/3JK45

Provenance

¹Contributed to nemar in BIDS format.

²Curated & ingested by the EEGDash catalog; see CITATION.cff for canonical reference.

³Persistent identifier: 10.17605/OSF.IO/3JK45.

Related & sibling datasets

NM000203EEG · 13 subj ON005262EEG · 12 subj NM000157EEG · 19 subj ON004018EEG · 16 subj ON007176EEG · 45 subj

+ 1 more — see See Also below →

BIDS

BIDS 1.9.0

Sidecars

not yet probed

Provenance

CC-BY 4.0 · 10.17605/OSF.IO/3JK45

Machine-readable

schema.org/Dataset · Croissant

Mirrors

OpenNeuro · NEMAR · HF org

Dataset ID	`NM000232`
Title	THINGS-EEG2: A large and rich EEG dataset for modeling human visual object recognition
Author (year)	`Gifford2019`
Canonical	—
Importable as	`NM000232`, `Gifford2019`
Year	2022
Authors	Alessandro T. Gifford, Kshitij Dwivedi, Gemma Roig, Radoslaw M. Cichy
License	CC-BY 4.0
Citation / DOI	doi:10.17605/OSF.IO/3JK45
Source links	OpenNeuro \| NeMAR \| Source URL

NM000232: eeg dataset, 10 subjects#

Quickstart#

About This Dataset#

Cohort#

Dataset Statistics#

Signal · Electrodes & live trace#

NEMAR Processing Statistics#

Manifest#

File Explorer#

API Reference#

Citation

Provenance

Related & sibling datasets

See Also#