Optimized feature mapping for eye movement recognition using electrooculogram signals
Harikrishna Mulam, Malini Mudigonda · 2017
This paper presents a novel method for eye movement recognition using Electrooculography (EOG) signals which are being used to control the Human-Computer Interface (HCI) systems. With the knowledge of several related investigations, this paper develops a methodology for eye movement recognition by introducing a feature mapping process. The proposed feature mapping process transforms the decomposed EOG signal into a transformation plane, where the intra-class margin is low. To accomplish the transformation, this proposed feature mapping process exploits Grey Wolf Optimization (GWO). The resultant features are used to classify the eye movements using Neural Network (NN). Later, the paper validates the superior performance of the proposed method with suitable performance analysis related to error minimization, mapping margin and recognition performance.