Abstract
Facial appearances like a happy, sad, disgust, surprise, angry, fear and Eye Gaze Estimations are the fastest means of communication while conveying any type of information. Expressions not only expose the sensitivity or feelings of any person but can also be used to judge his/her mental states. Therefore, it has become a wide interest area of research due to its applications to the fields like Human Computer Interface (HCI), Man Machine Interface (MMI) etc. In order to get more effective human behavior identification, we have presented a combination of facial expression recognition and eye gaze estimation technique where we have to recognize the real time expressions by using Convolutional Neural Network (CNN) model with transfer learning method. After that, we determine the eye gaze by using Viola Jones, Circular Hough Transform (CHT) and a geometric method. With the combination of both processes, we able to predict the human behavior like drivers concentration levels. In this way experiments are carried out on MUG database for our own trained CNN model based on VGG16 (Visual Geometry Group) pre trained model and gives better performance with accuracy of 93% for training and 94% for overall process.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call