Blink / Non-Blink Events Extracted from Video-EEG Data

Blink and non-blink events extracted from video-EEG data, with each event lasting one second. The dataset was authored by Valerii Zuev and last updated on May 11, 2026. It is shared under an MIT license and packaged in a 2.8 MB ZIP file.

Use Cases

• Train blink detection classifiers based on one-second video clips.
• Analyze correlations between eye-blink events and EEG signals.
• Benchmark machine learning models for temporal action recognition in video data.
• Study neurological or behavioral patterns through synchronized video and physiological data.

Strengths

• Dataset is shared under a permissive MIT license, allowing for broad reuse.
• A specific event duration of one second is defined for each clip.
• Data originates from synchronized video-EEG recordings, suggesting a multi-modal foundation.

Limitations

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

Provenance

Source

figshare

Collection Method

Extracted from video-EEG data.

Freshness

Last updated 2026-05-11 09:33:02; freshness should be verified.