摘要: We consider the problem of distributed face recognition in a calibrated camera sensor network. assume that each is given small and possibly different training set images taken under varying viewpoint, expression, illumination conditions. Each can estimate pose identity new using classical techniques such as Eigenfaces or Tensorfaces combined with simple classifier. However, estimates obtained by single could be very poor, due to limited computational resources, impoverished sets, etc., which lead poor results. Our key contribution propose algorithm neighboring cameras share their individual order achieve "consensus" on pose. For this purpose, we use provably convergent consensus SE(3) global Karcher mean fashion. Experiments Weizmann database show our algo- rithm effectively improves local estimates, achieves performance centralized algorithms only processing.