Recognition of human faces using anatomical knowledge

表达人类情感最有力的方式是面部表情。近年来，人们积极研究用户情感的检测，以最大限度地实现人机交互。特别是面部表情识别是人机交互（HCI）领域的一个活跃研究领域。在这个项目中，我们提出了一种从图像序列中识别面部表情的技术。该技术使用基于肌肉的面部模型来跟踪面部特征，如眉毛、眼睛和嘴巴。通过将模型与人脸图像进行匹配，得到面部肌肉的收缩向量。面部表情的似然度是在这些向量和主表情的代表向量之间定义的。实验结果表明，肌肉收缩向量能有效地反映面部表情的变化，并能识别出快乐、惊讶、愤怒和悲伤四种面部表情。

The most powerful way to express human emotion is the facial expression. Recently, detection of user's emotion have been actively studied to maximize human and computer interaction. In particular, recognition of facial expressions is an active research area in human and computer interaction (HCI).In this project, we have presented a technique for recognizing facial expressions from image sequences. The technique uses a muscle-based facial model for tracking facial features, such as eyebrows, eyes, and a mouth. The contractile vectors of facial muscles are obtained from the deformations of the model by matching the model with facial images. Plausibilities of facial expressions are defined between those vectors and representative vectors of principal expressions.Experimental results show that the muscular contractile vectors show the changes of facial expressions effectively and that four facial expressions, happiness, surprise, anger, and sadness, can be recognized by those vectors.