PROJECT TITLE :

A Novel Eye Localization Method with Rotation Invariance - 2014

ABSTRACT:

This paper presents a unique learning methodology for precise eye localization, a challenge to be solved in order to boost the performance of face processing algorithms. Few existing approaches will directly detect and localize eyes with arbitrary angels in predicted eye regions, face images, and original portraits at the identical time. To preserve rotation invariant property throughout the whole eye localization framework, a codebook of invariant native features is proposed for the illustration of eye patterns. A heat map is then generated by integrating a two-category sparse illustration classifier with a pyramid-like detecting and locating strategy to satisfy the task of discriminative classification and precise localization. Furthermore, a series of prior information is adopted to boost the localization precision and accuracy. Experimental results on 3 completely different databases show that our method is capable of effectively locating eyes in arbitrary rotation situations (360° in plane).