MER-PS: Physiological Signal-Based Emotion Recognition Data for Valence-Arousal Regression

MER-PS provides synchronized EEG and fNIRS recordings for continuous emotion modeling. The public training and validation data includes 24 anonymized subjects with dynamic labels for valence and arousal dimensions at 1 Hz. This data was released by MER-PS in 2026 for a Codabench regression task.

Use Cases

• Regression modeling of continuous emotional valence based on synchronized EEG and fNIRS signals.
• Regression modeling of continuous emotional arousal based on synchronized EEG and fNIRS signals.
• Developing multimodal fusion techniques for emotion recognition using physiological data streams.
• Benchmarking models on the Codabench valence-arousal regression task.

Strengths

• Data includes synchronized recordings from two physiological modalities: EEG and fNIRS.
• Dynamic emotional state labels are provided at a 1 Hz sampling rate for two affective dimensions.
• Subject identifiers are anonymized, covering 24 distinct individuals.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and data size are unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

MER-PS

Collection Method

Likely collected from laboratory recordings of physiological signals during emotional stimuli.

Freshness

Last updated 2026-04-28 12:31:26; freshness should be verified.