AI-Based Eye Tracking for Human-Computer Interaction

Human-computer interaction (HCI) is a field that has grown enormously in importance in recent times, with the spread of the personal computer. However, HCI devices have had little evolution since their first designs. Prolonged use of the mouse can cause injury or permanent damage to the hands and wr...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE International Conference on Electro Information Technology S. 241 - 246
Hauptverfasser: Crespo, David Sancho, Yu, Xinrui, Saniie, Jafar
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 30.05.2024
Schlagworte:
artificial intelligence
computer mouse
eye tracker
gaze vector
human-computer interaction
MediaPipe
Neural networks
Nonhomogeneous media
Nose
Tracking
Vectors
Webcams
Wrist
ISSN:2154-0373
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Human-computer interaction (HCI) is a field that has grown enormously in importance in recent times, with the spread of the personal computer. However, HCI devices have had little evolution since their first designs. Prolonged use of the mouse can cause injury or permanent damage to the hands and wrists. So, this paper proposes an alternative to the conventional mouse that does not require the use of hands and works only with a webcam. A program has been developed that performs all the functions of the mouse based on the image of the user's face captured by the camera. The cursor moves to the position of the screen where the user is looking at in real time, while the rest of the functionalities such as clicks and scrolling are controlled with facial commands, such as winking or tilting the head. To implement these functions, several tools based on artificial intelligence have been used, such as a facial landmark detector, a gaze vector estimator, and a fully connected multilayer neural network.
ISSN:2154-0373
DOI:10.1109/eIT60633.2024.10609928