A framework for real-time orientation detection.

A web-based face orientation recognition system allows virtual platforms to track the facial orientation of individuals accurately. This technology is valuable for organizations and educational institutions, enabling them to supervise employees and students during online meetings and classes effectively. In this research, the researchers employed the MediaPipe Facemesh solution to detect and map facial features and landmarks, allowing them to extract detailed facial geometry data. Utilizing this geometric data, the researchers calculated Euler rotational angles, which served as crucial indicators of facial orientation. By analyzing these Euler rotational angles, the team successfully recognized and determined the orientation of the face. The suggested approach achieves an 86.5% accuracy in detecting face orientation when the camera is positioned within a maximum distance of 4 feet from the subject's face. [ABSTRACT FROM AUTHOR]