A genuine smile is indeed in the eyes – The computer aided non-invasive analysis of the exact weight distribution of human smiles across the face. (October 2019)

• Record Type: Journal Article
• Authors: Ugail, Hassan; Al-dahoud, Ahmad
• Is Part Of: Advanced engineering informatics. Volume 42(2019)
• Publication Date: 2019-10
• Subjects: Smile analysis -- Genuine smiles -- Posed smiles -- Smile weight distribution; Computer-aided engineering -- Periodicals; Engineering -- Data processing -- Periodicals; 620.00285
• DOI: 10.1016/j.aei.2019.100967
• Languages: English
• ISSNs: 1474-0346

Abstract

Highlights: A lightweight computational framework to analyse the dynamics of human facial expressions is presented. This framework can undertake a detailed study of the smile expression. Using optical flow, we are able to accurately map the dynamic weight distribution of the smile expression. We show that the genuine smile is indeed in the eyes. Abstract: Understanding the detailed differences between posed and spontaneous smiles is an important topic with a range of applications such as in human-computer interaction, automatic facial emotion analysis and in awareness systems. During the past decade or so, there have been very promising solutions for accurate automatic recognition and detailed facial emotion analysis. To this end, many methods and techniques have been proposed for distinguishing between spontaneous and posed smiles. Our aim here is to go beyond the present state of the art in this field. Hence, in this work, we are concerned with understanding the exact distribution of a smile – both spontaneous and posed – across the face. To do this, we utilise a lightweight computational framework which we have developed to analyse the dynamics of human facial expressions. We utilise this framework to undertake a detailed study of the smile expression. Based on computing the optical flow across the face – especially across key parts of the face such as the mouth, the cheeks and around the eyes – we are able to accurately map the dynamic weight distribution of the smile expression. To validate our computational model, we utilise two publicly available datasets, namely the CK + dataset in which the subjects express posed smiles and the MUG dataset in which the subjects express genuine smiles. Our results not only confirm what already exists in the literature – i.e. that the spontaneous genuine smile is truly in the eyes – but it also gives further insight into the exact distribution of the smile across the face.