Driver monitoring systems (DMS) hold promise for improving road safety by detecting driver drowsiness and distraction, but what if these systems fall short? How do limitations in the algorithm’s capabilities to provide signals that are both robust and accurate affect the system’s ability to enhance traffic safety?

This talk will delve into:

• The shortcomings of existing driver monitoring systems
• Key algorithm approaches for extracting signals like gaze and eye openness
• The impact of algorithm choice on the rate of incorrect or missed warnings for distraction and drowsiness

• What is the definition of a good DMS from the driver perspective?
• What is the definition of a good DMS from the OEM perspective?
• What is the definition of a good DMS from the safety organizations perspective?
• What conditions should the DMS cover, and how do we handle performance degradation?
• What actions to take and when to inform the driver?