Estimating Face orientation from Robust Detection of Salient Facial Structures

This paper addresses the problem of estimating face orientation from automatic detection of salient facial structures using learned robust features. Face imagettes are detected using color and described using a weighted sum of locally normalized Gaussian receptive fields. Robust face features are learned by clustering the Gaussian derivative responses within a training set face imagettes. The most reliable clusters are identified and used as features for detecting salient facial structures. We have found that a single cluster is sufficient to provide a detector for salient facial structures that is robust to face orientation, illumination and identity. We describe how clusters are learned and which facial structures are detected. We show use of this detection to estimate facial orientation.