Exploiting the golden ratio on human faces for head-pose estimation

In this paper, a novel method for automatic head pose estimation is presented. This is based on a geometrical model of the head, in which basic features for estimating the pose consist in eyes and nose coordinates only. Worth noting, the majority of state-of-the-art approaches requires at least five features. The novelty of our work is the exploitation of the Vitruvian man’s proportions and the related “Golden Ratio”. The “Vitruvian man” is the well-known master-work by Leonardo Da Vinci, never used for automatic head pose estimation. Proposed method is compared by experiments with state-of-the-art ones, and shows a competitive performance although its simplicity and its low computational complexity.