Fusion of perceptual cues for robust tracking of head pose and position

The paradigm of perceptual fusion provides robust solutions to computer vis ion problems. By combining the outputs of multiple vision modules, the assu mptions and constraints of each module are factored out to result in a more robust system overall. The integration of different modules can be regarde d as a form of data fusion. To this end, we propose a framework for fusing different information sources through estimation of covariance from observa tions. The framework is demonstrated in a face and 3D pose tracking system that fuses similarity-to-prototypes measures and skin colour to track head pose and face position. The use of data fusion through covariance introduce s constraints that allow the tracker to robustly estimate head pose and tra ck face position simultaneously. (C) 2001 Pattern Recognition Society. Publ ished by Elsevier Science Ltd. All rights reserved.