Intelligent System to Help Autistic Children Recognize Emotions From: AlphaGalileo - 10/19/2009 Computer scientists from Nanyang Technological University in Singapore are developing an intelligent facial expression recognition system that would help autistic children understand the emotions of other people. The system makes use of derivative-based filtering to locate the edge of the human face, as well as boosting classifier to recognize facial expressions. Teik-Toe Teoh, Yok-Yen Nguwi, and Siu-Yeung Cho of the Center for Computational Intelligence in the School of Computer Engineering use Gaussian derivatives and Laplacian derivatives, and filters out non-face images using Adaboost. The technique is capable of finding key fiducial points for feature extraction and selection processing. Meaningful features are then classified into the corresponding classes. The team says the use of derivative filtering and boosting classifier makes for an efficient facial expression recognition system, which would be a portable device. Read the entire article at: http://www.alphagalileo.org/ViewItem.aspx?ItemId=62036&CultureCode=en