Multi-level smile intensity measuring based on mouth-corner features for happiness detection

Abstract

This work presents a low-complexity algorithm for multi-level smile intensity measurement based on mouth-corner features (MCFs). The proposed MCFs-based algorithm uses the mouth region images and accumulates the value of enhanced grayscale pixels along the horizontal axis. To further analyze the mouth shape, the local maximum information of the accumulated values is extracted to identify the height and width sizes of an opening mouth. Finally, the normalized threshold method is adopted to measure the smile intensity. The experimental results have shown that the proposed approach can achieve an average accuracy rate of 80% for the smiling measuring with four different levels of intensity. Moreover, the proposed algorithm can measure the smile intensity in a multi-face environment. Such results have demonstrated the efficiency and the feasibility of proposed algorithm.

Masked smiles matter – employee verbal expertise and emotion display during COVID-19

“…The manipulation of the FLE emotion display was tested with three items from Chang et al (2014;e.g. 'The FLE smiles.')…”