Estimation eye gaze direction is useful in various human-computer interactiontasks. Knowledge of gaze direction can give valuable information regardingusers point of attention. Certain patterns of eye movements known as eyeaccessing cues are reported to be related to the cognitive processes in thehuman brain. We propose a real-time framework for the classification of eyegaze direction and estimation of eye accessing cues. In the first stage, thealgorithm detects faces using a modified version of the Viola-Jones algorithm.A rough eye region is obtained using geometric relations and facial landmarks.The eye region obtained is used in the subsequent stage to classify the eyegaze direction. A convolutional neural network is employed in this work for theclassification of eye gaze direction. The proposed algorithm was tested on EyeChimera database and found to outperform state of the art methods. Thecomputational complexity of the algorithm is very less in the testing phase.The algorithm achieved an average frame rate of 24 fps in the desktopenvironment.
展开▼
