Title: Head Pose 3D Data Web-based Visualization2015,
Conference: Web3D ’15, June 18 – 21, 2015, HERAKLION, Greece, ACM 978-1-4503-3647-5/15/06.
Abstract
An approach build on discriminative random regression forests was followed in order to achieve fast, accurate and reliable estimation of head pose in uncontrolled environment. Data representing the changes of a person’s head direction, concerning two Degrees of Freedom (DOF), pitch and yaw, are collected and stored adopting a lightweight data exchange format (JavaScript Object Notation-JSON). After that, a web visualization approach is proposed in order to improve the understanding and the analysis of the captured 3D data.