Title :
Non-intrusive eye gaze estimation without knowledge of eye pose
Author :
Yoo, Dong Hyun ; Chung, Myung Jin
Author_Institution :
Dept. of Electr. Eng. & Comput. Sci., Korea Adv. Inst. of Sci. & Technol., Daejeon, South Korea
Abstract :
Eye gaze estimation is to detect the eye gaze of human. By detection of the eye gaze, a machine can read the intention. Eye gaze is very useful to read human intention. In this paper, a new eye gaze estimation method is suggested. The method allows large head movement without knowledge of eye pose. Also, feature detection is very important to achieve good performance. However, pupil detection is difficult because pupil color is similar to iris color. This paper proposes a robust feature detection method. This eye gaze estimation method is powerful, and it can work in real-time.
Keywords :
behavioural sciences; feature extraction; transforms; feature detection; human intention detection; nonintrusive eye gaze estimation; projective transform; Cameras; Charge coupled devices; Charge-coupled image sensors; Computer vision; Cornea; Humans; Light emitting diodes; Magnetic heads; Optical reflection; Robustness;
Conference_Titel :
Automatic Face and Gesture Recognition, 2004. Proceedings. Sixth IEEE International Conference on
Print_ISBN :
0-7695-2122-3
DOI :
10.1109/AFGR.2004.1301630