Multi-scale feature extraction for face recognition

Face recognition has been a very active research area in the past two decades. Many attempts have been made to understand the process how human beings perceive human faces. It is widely accepted that face recognition may rely on both componential cues (such as eyes, mouth, nose, cheeks) and non-componential/holistic information (the spatial relations between these features), though how these cues should be optimally integrated remains unclear. In this paper, we present a new different observer's view approach using multiscale feature extraction from face images. The basic idea of the proposed method is to construct facial features from multi-scale image patches from different face components and then employ a subspace PCA method for further dimensionality reduction and good representation of facial features. Finally, combining the contributions of each component features draws the recognition decision. 2,388 frontal face images of FERET face database are used for evaluating the proposed method and results are encouraging.