Deep Learning Approach for Driver Drowsiness Detection in Real Time
摘要
The paper suggests a deep learning method using convolutional neural networks (CNNs) for real-time driver drowsiness detection. A large number of accidents worldwide are caused by drowsiness, which presents a serious risk to road safety. With approximately 20% of drivers experiencing drowsiness at any given time, timely detection of microsleep and tiredness becomes paramount. The proposed CNN-based model categorizes driver characteristics into four classes: closed-eye expressions, open-eye expressions, yawns, and no yawns. Leveraging CNN, the model achieves close to 90% accuracy and is deployable on Android applications. This research presents a crucial step toward enhancing road safety by providing an efficient and accurate method for identifying driver drowsiness in real-time scenarios.