Efficient Performance Evaluation for Robust Eye Localization System

Accurate eye localization is very important and fundamental to all eye tracking applications. This task is challenging due to variation in appearance, size, shape, facial expression and lighting condition of an eye. Numerous methods have been developed to meet these challenges in past few years. To evaluate the performance of these methods precisely and to make a fair comparison of the accuracy of these methods, it is necessary to categorize them and evaluate them using standard Error Measure Metrics and standard Benchmark Database. We have assessed system architecture of eye localization model, various error measure metrics and benchmark databases to derive their appropriateness in evaluating performance of various methods. Keywords:- Gaze Tracking, Eye localization