Researchers are working towards a new drunk-driving detection AI software. Explore more here

The researchers at the Vietnam National University of Ho Chi Minh City are working towards a new drunk-driving detection AI software. This is to help with determining when a driver is intoxicated to avoid any impaired driving, which is a major concern on a global scale.

The network is different from other detectors introduced in the past, as indicated in the International Journal of Intelligent Information and Database Systems. The AI software is usually developed to focus on the person's eyes, the position of their head, and other functional state indicators. This new artificial intelligence actually posits thermal imaging which is a much less ambiguous way of indicating the intoxication state and allows a non-invasive scan of a driver's face.

The researchers also explain that earlier efforts at developing a way to detect drunkenness have focused on eye state, head position, or functional state indicators. However, such systems might be confused by other factors. The team points out that thermal imaging analysis offers a less ambiguous approach that is also non-invasive and could allow the authorities to screen people in city centers or at events where alcohol is likely to be consumed and people may opt to drive home.

Globally, the average adult consumes the equivalent of about one bottle of wine per week. That may seem a reasonable rate, but many people can't stop there. Serious health and social problems emerge around binge or excessive drinking: illness, crime, traffic accidents, addiction, etc. Alcohol is a factor in 5.3 percent of all deaths worldwide.

There is increasing attention on machine learning, deep learning, IoT, and computer vision technologies in attempts to reduce the damage done by alcohol and improve the safety of drinkers. AI-powered models can ensure alcohol purity, preventively monitor and assess human behaviors related to drinking, and generate support and services for addicted, intoxicated, or unconscious people.

Instead, the team behind the creation posits thermal imaging may be a less ambiguous way to tell if someone's drunk, and could one day allow police to non-invasively scan drivers' faces to ensure they don't get behind the wheel after downing too many beers.

One of the most important aspects of developing such an algorithm is its degree of accuracy. After all, as Tech Xplore pointed out, a false negative could have terrible consequences if it allows a drunk person to drive. In contrast, a false positive could hinder others from their commute without reason. At the moment, the network boasts an accuracy rate of 93%, though training the artificial intelligence on larger datasets or on a more diverse population of thermal images could help it become even more accurate in the future.

The team points out that it is important that any system designed to identify intoxicated people must have a very low rate of false positives and false negatives. After all, a false negative might see a drunk person driving their car whereas too many false positives would preclude sober drivers from using their vehicles and lead to frustration and a loss of trust in the system among the public.

There will always be a compromise in any such system, erring on the side of caution would be preferable, but optimizing the classification through larger training datasets on a diverse population of thermal images should bring it closer to the ideal, which would, of course, be the theoretically unachievable 100% accuracy with zero false positives, and zero false negatives.

The team points out that it is important that any system designed to identify intoxicated people must have a very low rate of false positives and false negatives. After all, a false negative might see a drunk person driving their car whereas too many false positives would preclude sober drivers from using their vehicles and lead to frustration and a loss of trust in the system among the public.