No abstract

Driver drowsiness is one of the reasons for a large number of road accidents in the world. In this paper, we have proposed an approach for the detection and prediction of the driver's drowsiness based on his facial features. This approach is based on deep learning techniques using convolutional neural networks CNN, with Transfer learning and Training from Scratch, to train a CNN model. A comparison between the two methods based on model size, accuracy and training time has also been made. The proposed algorithm uses the cascade object detector (Viola-Jones algorithm) for detecting and extracting the driver's face from images, the images extracted from the videos of the Real-Life Drowsiness Dataset RLDD will act as the dataset for training and testing the CNN model. The extracted model can achieve an accuracy of more than 96% and can be saved as a file and used to classify images as driver Drowsy or Non-Drowsy with the predicted label and probabilities for each class.

No abstract

No abstract

No abstract