Nexdata/500_Drivers_7_Expression_Recognition_Data

Dataset Overview

Description

This dataset includes seven facial expression recordings from 500 drivers, covering multiple age groups, time periods, and expressions. The capture devices were visible‑light and infrared binocular cameras. The dataset is suitable for driver facial expression analysis and other related tasks.

Data Size

• Number of subjects: 500

Demographics

• Gender: male, female
• Ethnicity: Vietnamese
• Age groups: 18–45, 46–60, over 60

Capture Environment

• Capture device: in‑vehicle camera

Data Diversity

• Expression diversity: multiple expressions
• Age diversity: multiple age groups
• Time diversity: multiple time periods

Equipment

• Camera type: visible‑light and infrared binocular cameras
• Resolution: 1920×1080

Capture Positions

• Center of the interior rear‑view mirror
• Above the central console
• Above the left A‑pillar
• Steering wheel area

Capture Times

• Day, dusk, night

Vehicle Types

• Sedan, SUV, MPV, truck, bus

Data Format

• Video format: .mp4

Accuracy

• Expression capture accuracy: >95%
• Label annotation accuracy: ≥95%

License Information

• License type: Commercial License