Abstract
Recently, eye tracking technology, with its special advantages of applying to such as ALS(amyotrophic lateral sclerosis) patients, is useful as a solution for disabilities or patients to interact with computer and others. In this paper, neural network has been employed in the calibration and tracking process for improving the accuracy and flexibility of the system. In order to verify the effect of the neural network, a human-computer interaction experiment of instruction recognition by human who is equipped a low-cost eye tracking device improved from HMD(Head-Mounted Display) has been performed. The experiment results show that the learning process of neural network can improve accuracy and flexibility of eye tracking system.
