AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |
Back to Blog
Eye of the beholder 3 setup12/29/2023 Over the last two decades, many techniques have been proposed to accurately estimate the gaze. Eye movements play a very significant role in human computer interaction (HCI) as they are natural and fast, and contain important cues for human cognitive state and visual attention. Figure 1: Human eye acting like a dipole and electrodes attached around it. This technique is not suitable for domestic or commercial use as it has to deal with close contact of electrodes near eye region but it is used for clinical purposes as well as in projects like MONEOG. One advantage of this method is that it can track the eye gaze even when the eye is closed. When the eye moves from center position towards one of those electrodes, the approached one experiences positive side while the other one experiences the negative side of the retina resulting in potential difference which helps in locating the eye's position. Electrodes are placed at the left and the right of the eye for horizontal tracking whereas above and below for vertical tracking. In this method, electrodes are attached to the skin around the eyes (see figure 1) that makes it possible to record the vertical and horizontal movements separately. II.METHODS OF EYE TRACKING Electro-Oculography is a method in which the potential between the back and the front of the human eye is measured as it acts like a dipole. The paper concludes in section 6 with the description of the possibilities of future work on eye and gaze tracking. Section 5 will cover category-divided tabular information of all the techniques used along with the models. Section 4 will cover the methods of gaze estimation and section 5 will be categorizing the models of gaze estimation where we review the techniques working on that particular model. In section 3, we will go through the categories of the eye models and review the techniques that use that particular model. In section 2, we will go through and understand the various methods of eye tracking. This research survey paper focuses on the methods and models of eye tracking and gaze estimation. It also helps us learn how a human reacts to a particular situation, for example, how we look at things and what we first see in a product or an advertisement so that we can enhance or change the old techniques to make them efficient and less time consuming. This not only helps the physically disabled people of our society but also provides a better path where one day we will be able to interact in a faster and a better sense. Eye tracking and gaze estimation opened many doors in the field of HCI because now we can use our eyes to interact with the computers. Today we live in the world of high processing speed and hence it is possible to enhance the way we interact with computers. All of this happens in fractions of a second. For example, even for simple events like a button click on a webpage, our brain first analyses the data from what we see and then sends the signal to our hand that is controlling the mouse at that particular moment. What we see next depends on what we saw initially while our face reactions, along with the gestures, are the outcome of every input our brain receives. In the latest technology, where we are moving towards human and computer interaction (HCI), the first sense is our sight. The importance of the eye cannot be scaled or compared. As the natural tendency of a human being is to learn, eyes become our best friend. Our brain analyses the data and responds by the means of gestures or language. We touch, we hear, we smell, we see and we feel. I.INTRODUCTION Human body is a house of sensors. Index Terms – Eye detection and tracking, methods and models of eye tracking, head pose and gaze estimation, research survey paper, visible spectrum. Be it the lack of accuracy due to the design of fovea or the tracking challenges in visible spectrum, this paper covers an in-depth research survey of the recent as well as the archaic methods and models used for eye tracking, head pose and gaze estimation. Even though we have achieved a significant progress in the last few decades, some domains are still facing the limitations. – Eye tracking and Gaze estimation are the most challenging areas in computer vision.
0 Comments
Read More
Leave a Reply. |