Robust Adapted Principal Component Analysis for Face Recognition

Abstract

Recognizing faces with uncontrolled pose, illumination, and expression is a challenging task due to the fact that features insensitive to one variation may be highly sensitive to the other variations. Existing techniques dealing with just one of these variations are very often unable to cope with the other variations. The problem is even more difficult in applications where only one gallery image per person is available. In this paper, we describe a recognition method, Adapted Principal Component Analysis (APCA), that can simultaneously deal with large variations in both illumination and facial expression using only a single gallery image per person. We have now extended this method to handle head pose variations in two steps. The first step is to apply an Active Appearance Model (AAM) to the non-frontal face image to construct a synthesized frontal face image. The second is to use APCA for classification robust to lighting and pose. The proposed technique is evaluated on three public face databases — Asian Face, Yale Face, and FERET Database — with images under different lighting conditions, facial expressions, and head poses. Experimental results show that our method performs much better than other recognition methods including PCA, FLD, PRM and LTP. More specifically, we show that by using AAM for frontal face synthesis from high pose angle faces, the recognition rate of our APCA method increases by up to a factor of 4.

Illumination invariant face recognition: A survey of passive methods

“…The problem is even more difficult in applications where only one gallery image per person is available. Chen, Lovell and Shan [8] describe a recognition method, Adapted Principal Component Analysis (APCA), which can simultaneously deal with large variations in both illumination and facial expressions using only a single gallery image per person. Experimental results show that APCA performs much better than other recognition methods including PCA and LDA.…”

DISGUISED DISCRIMINATION OF LOCALITY-BASED UNSUPERVISED DIMENSIONALITY REDUCTION

“…Unsupervised DR is an important and necessary branch of DR techniques as collections of plentiful unlabeled samples are easy but achievements of class information are relatively expensive. 31 Principal Component Analysis (PCA) 8,15 is one of the most popular unsupervised DR algorithms. It seeks a set of orthogonal projection vectors that best represents the data in the least-squares sense by maximizing the scatter of data.…”

“…It has been demonstrated that the locality-based LPP can usually preserve the manifold structure in the embedding space to a certain extent, and thus is relatively e®ective on some classi¯cation tasks (e.g. face recognition 8,13 ). As a result, the \locality" idea is constantly generalized, and thus many locality-based dimensionality reduction (DR) algorithms have successively been proposed, typically including unsupervised Unsupervised Discriminant Projection (UDP) a29 and OLPP, 6 semi-supervised discriminant analysis 5,26 and supervised MFA, 24 LDE, 7 ANMM, 23 GSSGE, 30 etc.…”