Aggregating Partial Least Squares Models for Open-set Face Identification

Abstract

Face identification is an important task in computer vision and has a myriad of applications, such as in surveillance, computer forensics and human-computer interaction. In the past few years, several methods have been proposed to solve face identification task in closed-set scenarios. Most of them make assumption of the complete knowledge of the world. However, in real-world applications, one might want to determine the identity of an unknown face, that is, a face whose identity does not match any known individual, comprising the open-set scenario. In this work, we propose a novel method to perform open-set face identification by aggregating Partial Least Squares models in a simple but fast way. Evaluation is performed in four datasets: FRGCv1, FG-NET, Pubfig and Pubfig83. Results show significant improvement when compared to state-of-the art approaches regardless challenges posed by different datasets.