Novel Method for Detection of Stress in Employees using Hybrid Deep learning Models

Currently, the employee well-being is considered to be one of the most significant concern. The corporation has been increasingly concerned with the health and happiness of its workers. The staff also have been pursuing different ways to improve their current well-being state. Yet, the well-being is frequently tied to the everyday activity and behavior, especially in the job where it influences stress and mood level. In other words, the quality of a person’s well-being is affected by the behavior at a job. In this proposed uses Neural Network approaches to investigate the causes of stress in the lives of working adults and identify the most important contributing variables. Upon receiving an input image, the following step is preprocessing, which in this proposed approach to entails image enhancement and the use of ROI to detect the mouth and eye regions. To extract the characteristics Gabor filter is employed. Eventually training the model using R-CNN. The R-CNN framework, which is faster, is employed in many contexts. This study's primary contribution is a multi-task loss function, called center loss, for use with the Fast R-CNN model to control the development of discriminative deep features for face/non-face classification. A second contribution of this research is the employment of a multi-scale training technique to enhance detection performance. The empirical findings of our study demonstrate the efficacy of the proposed method in detecting stress among IT workers because of its integration of the aforementioned strategies. With an accuracy of roughly 89.7 percent, this model performs better than the other two models, which are the convolutional neural network (CNN) and the recurrent neural network (RNN).