Fast Estimation of Gaussian Mixture Models for Image Segmentation 利用高斯混合模型实现图像分割TheExpectation-Maximizationalgorithmhasbeenclassicallyusedtofindthemaximumlikelihoodestimatesofparametersinprobabilisticmodelswithunobse
Towards a Practical Face Recognition System Many classic and contemporary face recognition algorithms work well on public data sets, but degrade sharply when they are used in a real recognition