A new method for extracting eye motion in real-time using deep learning
Paper ID : 1131-MVIP2020
Suhair Salehi *, Mohammad Hosein Shakoor, Reza Ghasemi
Many methods have been proposed to extract eye movement and gaze detection. One of the most important methods is deep learning algorithm which is a subset of machine learning. Due to the feature extraction with high accuracy it has attracted many experts. One of the most important and fastest methods in object detection is YOLO architecture based method. It is a real-time intelligent object recognition network. The YOLONet looks one at the image but in a clever way. This paper uses a new network called SuNet, the proposed method is similar to the YOLONet but it differs in layers.The proposed method takes up less processor space than the YoloNet. As a result it eliminates the need for a lower processor. It also produces better results in terms of speed and accuracy in detecting eye movement and depth of vision for moving an electric wheelchair by changing the layers of the YOLONet. This method, while having all the positive features of the existing methods, has a higher volume, speed and accuracy than those.
eye extraction,gaze detection,deep learning, machine learning, wheelchair
Status : Paper Accepted (Poster Presentation)