This post is also available in:
עברית (Hebrew)
A new study from Sharjah University has unveiled an innovative machine learning (ML) and deep learning (DL) algorithm designed to detect driver drowsiness, potentially saving thousands of lives each year. The research, published in Biomedical Signal Processing Control, offers a promising solution to a serious road safety issue—driver fatigue.
Professor Saad Harous, the lead researcher, notes that drowsiness is a leading cause of road accidents, with the U.S. National Highway Traffic Safety Administration estimating that it leads to 100,000 accidents annually, resulting in 1,500 deaths and 70,000 injuries. “Detecting driver drowsiness [has] become an important task that necessitates an automated system to detect and prevent these adverse outcomes early on,” Harous said, according to TechXplore.
The study uses electroencephalography (EEG), a widely recognized method for detecting sleep and drowsiness, and combines it with cutting-edge ML techniques, specifically convolutional neural networks (CNNs). CNNs, a type of deep learning algorithm, are adept at analyzing complex data, like brain signals. The researchers applied a random search optimization algorithm to fine-tune EEG preprocessing parameters, which has never been done before in drowsiness detection research.
After optimizing the parameters, the researchers achieved remarkable results. The accuracy of their system surged from 91% to 95%, and further fine-tuning using the Optuna Hyperparameter framework boosted the accuracy to 97%. The team’s CNN-SVM classifier, a combination of CNN with machine learning classifiers, reached an impressive 99.9% accuracy, while reducing the training time significantly.
This algorithm could soon help detect drowsiness with high precision, alerting drivers before fatigue leads to accidents. If adopted by transportation authorities, this technology can have a huge impact on society.
Despite the promising results, the researchers are still seeking industry interest for implementation. Harous hopes that the technology, potentially integrated into in-car systems, could become a game-changer in road safety. With this breakthrough, the team has taken a significant step toward reducing accidents caused by drowsy driving and improving public safety on the roads.
