A Simple Method for Facial Pose Detection(Image and Signal Processing)(<Special Section>Information Theory and Its Applications)

概要

論文の詳細を見る
Much of the work on faces in computer vision has been focused on face recognition or facial expression analysis, but has not been directly related with face direction detection. In this paper, we propose a vision-based approach to detect a face direction from a single monocular view of a face by using a facial feature called facial triangle, which is formed by two eyebrows and the lower lip. Specifically, the proposed method introduces simple formulas to detect face rotation, horizontally and vertically, using the facial triangle. It makes no assumption about the structure of the face and produces an accurate estimate of face direction.
社団法人電子情報通信学会の論文
2004-10-01