Face Analysis Under Uncontrolled Conditions

7Micro- and Macro-Expression Analysis

Benjamin ALLAERT¹, Ioan Marius BILASCO² and Chaabane DJERABA²

¹IMT Nord Europe, Lille, France

²University of Lille, France

7.1. Introduction

In this chapter, we focus our attention on the recognition of expressions in the presence of a great diversity of facial movement amplitudes. Therefore, we are really interested in the study of macro- and micro-expressions which allow us to provide an exhaustive panel of facial movements. As a reminder, macro-expressions are defined by voluntary movements and characterized by high-intensity movements lasting between 0.5 and 4 seconds. As for micro-expressions, they are often involuntary and last a fraction of a second, on average between 170 ms and 500 ms (Yan et al. 2013). Although these movements are very fast and generally not perceptible to the human eye, micro-expressions provide valuable information on the emotional state of a person.

A wide variety of approaches have been proposed for the recognition of macro-expressions in video sequences. However, these solutions are not suitable in the presence of low-intensity expressions such as micro-expressions (Yan et al. 2013). Although recent approaches tend towards common techniques for analyzing macro- and micro-expressions, there is no unified solution to optimally process both simultaneously. This is mainly due to the different characteristics of motion intensity change and/or texture.

In the previous chapter, we introduced a new descriptor, called ...