作者: Jingying Chen , Oliver Lemon
DOI: 10.1007/978-3-642-10520-3_56
关键词:
摘要: A robust and efficient facial feature detection tracking approach for head pose estimation is presented in this paper. Six points (inner eye corners, nostrils mouth corners) are detected tracked using multiple cues including intensity its probability distribution based on a novel histogram entropy analysis, geometric characteristics motion information. The estimated from 3D model POSIT RANSAC algorithms. proposed method demonstrates capability gaze new multimodal technology enhanced learning (TEL) environment supporting of social communication skills.