An Eye State Recognition System Using Transfer Learning: AlexNet-Based Deep Convolutional Neural Network
Ismail Kayadibi,Gür Emre Güraksın,Uçman Ergün,Nurgül Özmen Süzme
DOI: https://doi.org/10.1007/s44196-022-00108-2
IF: 2.259
2022-07-23
International Journal of Computational Intelligence Systems
Abstract:For eye state recognition (closed or open), a mechanism based on deep convolutional neural network (DCNN) using the Zhejiang University (ZJU) and Closed Eyes in the Wild (CEW) dataset, has been proposed in this paper. In instances where blinking is consequential, eye state recognition plays a critical part for the development of human–machine interaction (HMI) solutions. To accomplish this objective, pre-trained CNN architectures on ImageNet were first trained on the both dataset, which included both open and closed-eye states, and then they were tested, and their performance was quantified. The AlexNet design has proven to be more successful owing to these assessments. The ZJU and CEW datasets were leveraged to train the DCNN architecture, which was constructed employing AlexNet modifications for performance enhancement. On the both datasets, the suggested DCNN architecture was tested for performance. The achieved DCNN design was found to have 97.32% accuracy, 95.37% sensitivity, 97.97% specificity, 93.99% precision, 94.67% F 1 score, and 99.37% AUC values in the ZJU dataset, while it was found to have 97.93% accuracy, 98.74% sensitivity, 97.15% specificity, 97.11% precision, 97.92% F 1 score, and 99.69% AUC values in the CEW dataset. Accordingly, when compared to CNN architectures, it scored the maximum performance. At the same time, the DCNN architecture proposed on the ZJU and CEW datasets has been confirmed to be an acceptable and productive solution for eye state recognition depending on the outcomes compared to the studies in the literature. This method may contribute to the development of HMI systems by adding to the literature on eye state recognition.