Artificial intelligence system for driver distraction by stacked deep learning classification

Increasing efforts in the transportation system have recently improved driver safety and reduced crash rates. Lack of attention fatigue directly affect driver's consciousness. Driver distraction is an essential driver-specific factor practical applicability forward collision warning (FCW). However, there are still too many false alarms generated by existing FCW to be mitigated. This paper proposes facial detection identify features test anomalies' prediction against drivers using stacked convolutional neural network (CNN) layers. The proposed model used overlapping HAAR CNN classifications eye areas, such as open or closed. In addition sliding query window's overall intensity information. conventional function, which elevates brightness nearby regions, preferable. method considers current intelligent system-based solutions minimize effects continuously comparing with flexible thresholds. experimental results analyzed from accurate driving datasets. At 456 iterations, acquired over 80% accuracy, while loss near zero. implication risk tolerance further explored this manner. Several risks connected any type system.