EEG Recordings of Human Emotive Responses to Robot Facial Expressions

240 EEG trials from 10 healthy participants reacting to four facial expressions displayed by the Miko social robot. The dataset includes raw and cleaned EEG data recorded at 128 Hz, along with stimulus onset markers. It was created by researchers at Imperial College London and published in 2021.

Use Cases

• Training classifiers for emotion recognition based on EEG signals.
• Studying the neural correlates of human responses to robot facial expressions.
• Benchmarking signal processing techniques for artefact removal in EEG data.

Strengths

• Data from 10 participants provides a basis for individual and group-level analysis.
• Includes both raw and cleaned (ICA-processed) EEG versions for methodological comparison.
• Stimulus presentation was randomized, which helps control for order effects.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The participant pool of 10 healthy individuals is relatively small.

Provenance

Source

Imperial College London

Collection Method

EEG recorded from 16 unipolar channels during visual interaction with a social robot.