EEGdash›NeMAR›NM000112

Iss. 112 · 123 subjects · 123 recordings · CC-BY-4.0

Dataset Brief · FACED - Finer-grained Affective Computing EEG Dataset

NM000112: eeg dataset, 123 subjects#

FACED - Finer-grained Affective Computing EEG Dataset

Citation: Yisi Liu, Olga Sourina, Minh Khoa Nguyen (2023). FACED - Finer-grained Affective Computing EEG Dataset. 10.82901/nemar.nm000112

123-participant EEG dataset — FACED - Finer-grained Affective Computing EEG Dataset.

Data & curation Yisi Liu · Olga Sourina · Minh Khoa Nguyen
Year 2023 · Distributed via NeMAR

EEG · 32 ch250, 1000 HzBIDS 1.7.0Task · watchingVideoClips

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 NM000112

dataset = NM000112(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 = NM000112(cache_dir="./data", subject="01")

Advanced query

dataset = NM000112(
    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{nm000112,
  title = {FACED - Finer-grained Affective Computing EEG Dataset},
  author = {Yisi Liu and Olga Sourina and Minh Khoa Nguyen},
  doi = {10.82901/nemar.nm000112},
  url = {https://doi.org/10.82901/nemar.nm000112},
}

§ 02Study · The README

About This Dataset#

The Finer-grained Affective Computing EEG Dataset (FACED) contains scalp EEG recordings from 123 healthy participants who watched 28 emotion-eliciting video clips designed to evoke nine different emotion categories. The dataset includes four negative emotions (anger, fear, disgust, sadness) from Ekman’s basic emotions and four positive emotions (amusement, inspiration, joy, tenderness) selected based on recent psychological and neuroscience progress and application needs. Participants provided detailed self-reported emotion ratings on 12 dimensions: eight emotions, arousal, valence, liking, and familiarity. The dataset is designed to facilitate cross-subject affective computing research and development of EEG-based emotion recognition algorithms for real-world applications.

Participants (123 subjects, 75 female, ages 17-38, mean=23.2 years) were seated 60 cm from a 22-inch LCD monitor in a regular office environment. Each trial consisted of: (1) a 5-second fixation cross, (2) a video clip of varying length (typically 30-60 seconds), and (3) subjective emotional rating on 12 items (anger, fear, disgust, sadness, amusement, inspiration, joy, tenderness, valence, arousal, liking, familiarity) on a continuous 0-7 scale, followed by at least 30 seconds rest. Video clips were presented in blocks: three positive blocks, three negative blocks, and one neutral block, with 20 arithmetic problems between blocks to minimize carryover effects. The 28 video clips were designed to target nine emotion categories, with randomized presentation order across participants. EEG was recorded using a 32-channel biosignal recording system sampled at either 1000 Hz (92 subjects) or 250 Hz (31 subjects), with channels positioned according to the International 10-20 system. Signal units were recorded in either Volts or microVolts depending on the hardware configuration used.

Video stimulus information: The dataset includes 28 video clips designed to elicit nine emotion categories (Trigger values 1–28): - Anger (Videos 1-3): Durations 73-81 seconds, negative valence - Disgust (Videos 4-6): Durations 69-91 seconds, negative valence - Fear (Videos 7-9): Durations 56-106 seconds, negative valence - Sadness (Videos 10-12): Durations 45-82 seconds, negative valence - Neutral (Videos 13-16): Durations 35-43 seconds, neutral valence - Amusement (Videos 17-19): Durations 56-73 seconds, positive valence - Inspiration (Videos 20-22): Durations 76-129 seconds, positive valence - Joy (Videos 23-25): Durations 34-68 seconds, positive valence - Tenderness (Videos 26-28): Durations 54-77 seconds, positive valence

DOI

FACED - Finer-grained Affective Computing EEG Dataset

Introduction

Metadata for each video (duration, source film, source database, valence, targeted emotion) is read from Stimuli_info.xlsx. Event markers (from evt.bdf annotations): - 100: Task/block start

View full README

DOI

FACED - Finer-grained Affective Computing EEG Dataset

Introduction

Metadata for each video (duration, source film, source database, valence, targeted emotion) is read from Stimuli_info.xlsx. Event markers (from evt.bdf annotations): - 100: Task/block start - 101: Video onset - 102: Video offset - 1–28: Video index (appears just before 101, used to link to stimulus metadata) - 201/202: Block boundary markers - “Start Impedance” / “Stop Impedance”: Technical markers (ignored)

The conversion script reads evt.bdf annotations for each subject, parses video presentation spans (from video index + 101 to 102), and creates MNE Annotations with the source film title (video_title) as description. These annotations are exported to BIDS events.tsv with extra columns: - emotion_label: targeted emotion category (Anger, Disgust, Fear, Sadness, Neutral, Amusement, Inspiration, Joy, Tenderness) - binary_label: positive/negative/neutral classification - video_index: 1–28 - Self-reported ratings (Joy, Tenderness, Inspiration, Amusement, Anger, Disgust, Fear, Sadness, Arousal, Valence, Familiarity, Liking)

Description of the preprocessing if any

Raw BDF files from the biosignal recording system have been converted to BIDS format. Channel names are standardized to match the International 10-20 nomenclature. Subjects have been assigned numeric IDs (sub-000 through sub-122) corresponding to their original subject designations in the dataset. Recording dates have been set to a default value (2023-01-01) due to privacy considerations, while time relationships between files are preserved. Subject demographic information (age, sex) has been extracted from the Recording_info.csv file and properly formatted for BIDS.

Stimulus timing information from the evt.bdf event files has been parsed and enriched with metadata from Stimuli_info.xlsx. Each video presentation is annotated with the targeted emotion category (Anger, Disgust, Fear, Sadness, Neutral, Amusement, Inspiration, Joy, Tenderness) and includes self-reported ratings from After_remarks.mat when available.

Citation

When using this dataset, please cite: 1. Liu, Y., Sourina, O., & Nguyen, M. K. (2023). Finer-grained Affective Computing EEG Dataset. Scientific Data, 10(1), 809. https://doi.org/10.1038/s41597-023-02650-w 2. Synapse Platform: https://www.synapse.org/#!Synapse:syn50614194 3. The dataset is available at the Synapse platform repository.

Data curators: Pierre Guetschel (BIDS conversion) Original data collection team: - Yisi Liu (Nanyang Technological University) - Olga Sourina (Nanyang Technological University)

- Minh Khoa Nguyen (Nanyang Technological University)

Automatic report

Report automatically generated by ``mne_bids.make_report()``.

The FACED - Finer-grained Affective Computing EEG Dataset dataset was created

by Yisi Liu, Olga Sourina, and Minh Khoa Nguyen and conforms to BIDS version 1.7.0. This report was generated with MNE-BIDS (https://doi.org/10.21105/joss.01896). The dataset consists of 123 participants (comprised of 48 male and 75 female participants; handedness were all unknown; ages ranged from 17.0 to 38.0 (mean = 22.94, std = 4.66)) . Data was recorded using an EEG system (Biosemi) sampled at 1000.0, and 250.0 Hz with line noise at n/a Hz. There were 123 scans in total. Recording durations ranged from 3468.0 to 6743.0 seconds (mean = 4544.83, std = 647.24), for a total of 559013.71 seconds of data recorded over all scans. For each dataset, there were on average 32.0 (std = 0.0) recording channels per scan, out of which 32.0 (std = 0.0) were used in analysis (0.0 +/- 0.0 were removed from analysis).

§ 03Cohort · Participants

Cohort#

Dataset Statistics#

Age distribution by gender (n=123, range 17–38 yr, mean 22.9 yr)

1520253035

Female · 75Male · 48

Sex composition

Female
75
Male
48
F : M ratio
1.56 : 1

61% female · n = 123 subjects with reported sex.

Channel counts: 32 ch (n=123 recordings)

Sampling frequencies (Hz)

2501000

Total recording duration: 155 h

§ 04Signal · Electrodes & trace

Signal · Electrodes & live trace#

Fig. 01 Signal & montage 32 ch · EEG · 250, 1000 Hz · 123 subjects, 123 recordings

Live trace viewer — sub-021 · task-watchingVideoClips

Showing one representative recording out of 123 subjects and 123 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 · 30 sensors — 30 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.NM000112(cache_dir, query=None, s3_bucket=None, **kwargs)

Bases: EEGDashDataset

Author (year)Liu2024_112

Canonical—

Importable asNM000112 · Liu2024_112

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

class eegdash.dataset.NM000112(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

FACED - Finer-grained Affective Computing EEG Dataset

Study:: nm000112 (NeMAR)
Author (year):: Liu2024_112
Canonical:: —

Also importable as: NM000112, Liu2024_112.

Modality: eeg. Subjects: 123; recordings: 123; tasks: 1.

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/nm000112 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000112 DOI: https://doi.org/10.82901/nemar.nm000112

Examples

>>> from eegdash.dataset import NM000112
>>> dataset = NM000112(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 — NM000112.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 nm000112 to reproduce the tutorial on this dataset.

Citation

Yisi Liu, Olga Sourina, Minh Khoa Nguyen (2023). FACED - Finer-grained Affective Computing EEG Dataset. 10.82901/nemar.nm000112

Provenance

¹Contributed to nemar in BIDS format.

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

³Persistent identifier: 10.82901/nemar.nm000112.

Related & sibling datasets

ON007181EEG · 59 subj ON005628EEG · 102 subj ON003825EEG · 50 subj ON007827EEG · 99 subj ON007823EEG · 169 subj

+ 1 more — see See Also below →

BIDS

BIDS 1.7.0

Sidecars

events · events.json · channels · eeg.json

Provenance

CC-BY-4.0 · 10.82901/nemar.nm000112

Machine-readable

schema.org/Dataset · Croissant

Mirrors

OpenNeuro · NEMAR · HF org

Dataset ID	`NM000112`
Title	FACED - Finer-grained Affective Computing EEG Dataset
Author (year)	`Liu2024_112`
Canonical	—
Importable as	`NM000112`, `Liu2024_112`
Year	2023
Authors	Yisi Liu, Olga Sourina, Minh Khoa Nguyen
License	CC-BY-4.0
Citation / DOI	10.82901/nemar.nm000112
Source links	OpenNeuro \| NeMAR \| Source URL

NM000112: eeg dataset, 123 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#