Australian researchers develop AI to detect dangerous driving through facial analysis

Australian researchers have developed artificial intelligence (AI) technology that could change how drunk and dangerous drivers are identified, offering a non-invasive alternative to traditional breathalyzers.

Using 3D facial analysis, researchers are able to detect three major causes of road accidents simultaneously: blood alcohol concentration, fatigue and expressions such as anger, said a statement from Australia's Edith Cowan University (ECU) on Wednesday.

The system, developed by ECU researchers, uses a single deep learning model to identify blood alcohol concentration with nearly 90% accuracy and drowsiness with 95% accuracy, the statement said.

The technology can also determine the level of intoxication, classifying alcohol impairment into three categories – sober, moderate and severe, it said.

Lead researcher, ECU PhD candidate Abdullah Tariq, said the system automatically captures diverse facial dynamics, including eye blinking, subtle facial movements, and progressive facial feature changes that are critical for distinguishing between different states.

Unlike breathalysers, the method allows continuous, real-time monitoring without driver cooperation, the researchers said.

A companion study demonstrated that combining infrared and color video improves detection in low-light conditions, they said, adding the research could lead to an innovative, non-invasive way to help combat drunk driving.