Fast head pose estimation for human-computer interaction

Abstract

This paper describes a Hough Forest based approach for fast head pose estimation in RGB images. The system has been designed for Human-Computer Interaction (HCI), in a way that with just a simple web-cam, our solution is able to detect the head and simultaneously estimate its pose. We leverage the Hough Forest with Probabilistic Locally Enhanced Voting model, and integrate it into a system with a skin detection step and a tracking filter for the head orientation. Our implementation drastically speeds up the head pose estimations, improving their accuracy with respect to the original model. We present extensive experiments on a publicly available and challenging dataset, where our approach outperforms the state-of-the-art.