A Survey on Gaze Estimation

Eye orientation contains abundance of significant information for estimating user's intention and purpose. Large number of research on gaze direction has been given and improved in statistics, biometrics, organizational behavior, etc. Details of gaze methods are essential conditions for human-computer interactive control, human intention prediction, expression analysis, cognitive state analysis, etc. Although active and significant progress in last decades, eye detection still remains challenging to deviation of location, variability in scale, individuality of eyes, complexity of cameral, calibration of person and related instruments, fixation of head pose. In this paper, gaze estimation has been discussed and various models of gaze detection and tracking have been presented, which show current theoretical need to be further modified in order to improve general gaze tracking estimation solutions in computer vision and beyond.