Neural Network Identifies What Drivers Feel Before, During and After Maneuvers
Researchers at the St. Petersburg Federal Research Center of the Russian Academy of Sciences and ITMO University have developed a method for assessing drivers’ emotional profiles under real-world road conditions.
The scientists used an open dataset containing more than 300 hours of video and telemetry data from 42 drivers. Based on facial expressions and physiological indicators, the neural network identified six basic emotions and a neutral state.
The researchers assessed emotions in three phases: before, during and after a maneuver. This revealed several patterns: negative emotions intensified before a maneuver, peaked during the maneuver itself, and either subsided afterward or remained elevated. An individual emotional profile was created for each driver.
According to Kashevnik, the approach could eventually be used in adaptive driver-assistance systems, driver training and digital assistants that plan routes designed to minimize emotional stress.
