Facial Analysis Based Ai and Deep Learning Assisted Driver Sleep Detection Technique

The public's general safety is seriously threatened by the high frequency of traffic accidents brought on by tired drivers. The detection of driver sleepiness has advanced significantly in recent years due to the considerable confluence of artificial intelligence (AI) and intelligent cameras. By tracking facial expressions including yawning, eye and head movements, and other facial emotions, developments in computer vision technology enable the detection of driver sleepiness. Together with evaluations of the driver's behaviour and physical state, these physical indicators help detect weariness and reduce the risk of accidents involving sleepy drivers. This research offers a wide range of painstakingly created algorithms that were carefully examined to see how well they detected tiredness. The detection of sleepiness was then made easier using a variety of deep learning models, including MobileNet, VGG16, and a conventional CNN. The EfficientNet-B3 and Inception-ResNet-v2 model proved to be wise selection among these strategies for sleepiness detection in drivers because of its capacity to process data in real-time and adapt to resource-constrained conditions; the greatest accuracy of 97 % was attained.