Driver drowsiness detection using Convolutional Neural Networks-inspired features and Principal component analysis with K-Nearest Neighbors

Panganai, Marvelous Alexander; Nyandoro, Leslie Kudzai; Zvarevashe, Kudakwashe

doi:10.1109/zcict55726.2022.10045859

Cited by 1 publication

(1 citation statement)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A variety of works are done for drowsiness detection using machine learning. A study by M A Panganai et al [2] introduces two feature extraction techniques, Multilayer Perceptron (MLP) and Convolutional Neural Network (CNN), and utilizes Principal Component Analysis (PCA) for dimensionality reduction. These methods are combined with five classifiers, including Linear Discriminant Analysis (LDA), XGBoost, Logistic Regression (LR), Decision Tree, and K-Nearest Neighbours (KNN), to detect drowsiness in drivers.…”

Section: Introductionmentioning

confidence: 99%

A Deep Learning model for driver drowsiness detection in extremely low-light conditions

Anoop,

Deivanathan

2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

Accidents occurring at night account for a significant percentage when compared to those that happen during the daytime. One of the leading causes of such accidents is driver drowsiness, which could result from inadequate sleep or exhaustion. Continuous monitoring of the facial expressions of the driver can detect signs of drowsiness, and an alarm system can alert the driver and prevent potential accidents. Various research studies have been conducted in this area. However, the low lighting conditions at night make it challenging to obtain detailed images, making it difficult to apply the proposed models effectively. This paper presents an effective model for detecting driver drowsiness using deep learning techniques. The study utilizes AlexNet as a classifier and attains an impressive accuracy of 98.3% after hyperparameter tuning.

show abstract

Section: Introductionmentioning

confidence: 99%