EmoRoad Multimodal Dataset of Driver Responses Under Controlled Access

EmoRoad is a multimodal dataset containing psychological, physiological, and behavioral data from human subjects in diverse driving scenarios. The dataset includes identifiable facial video recordings, EEG signals, and other measures, requiring a Data Usage Agreement for access. It was created by Stephen Jia Wang and made available in 2026.

Use Cases

• Classify driver emotional states by fusing facial video features with EEG signal patterns.
• Analyze correlations between physiological signals (e.g., EEG) and behavioral measures during specific driving events.
• Train multimodal time-series models to predict psychological responses from concurrent video and physiological data streams.
• Benchmark privacy-preserving analysis techniques on identifiable facial video data under controlled access conditions.

Strengths

• Multimodal data collection includes video, EEG, and behavioral measures for holistic analysis.
• Access is governed by a formal Data Usage Agreement ensuring ethical use and participant protection.
• Dataset is citable via a DOI and was published on a reputable research data platform (figshare).

Limitations

• Dataset size and participant count are unspecified, limiting assessment of statistical power.
• Controlled access model and DUA requirements create a barrier for rapid prototyping and use.
• The raw data contains identifiable information (facial video), restricting sharing and requiring secure handling.

Provenance

Source

Stephen Jia Wang, The Hong Kong Polytechnic University.

Collection Method

Human-subject data collection in driving scenarios.

Time Range

null

Freshness

Last updated in April 2026.

Geography

null