Perspectives for the use of facial expressions

Facial Expressions

The human face (facial expressions) consists of a vast panorama in constant movement of what is happening in our most intimate state of mind, with a considerable diversity of nuances and great complexity.

Humanity has always been developing, in a natural way, the ability to understand the facial expressions of others, initially as a survival strategy (friend / enemy), and currently aiming at relationships of all kinds.

Based only on sensitivity and intuition, we all spend a considerable amount of time interpreting, analyzing and reacting to the perceptible signs of facial expressions of bosses, spouses, teachers, partners, co-workers, strangers, relatives. , etc.

With the technology and the capacity of data processing increasing, the face of each one of us and the information that is transmitted spontaneously for some time has ceased to be just the way we present, relate or behave in public.

Far from merely signaling our momentary state of mind or our reaction to situations, people or facts, this whole set of data has been increasingly treated as something strategic and object of intense studies.

The use of facial expression analysis continues to grow, whether with Apple using it to unlock the i-phone, in churches in the United States trying to attract believers, in England identifying those responsible for shoplifting, with the police in Wales arresting suspects at football games in China identifying undisciplined drivers, allowing tourists access to certain attractions, among others.

In the medical field, some applications are very promising, such as the early diagnosis of genetic diseases such as Hajdu-Cheney syndrome, in hopeful attempts to treat autism, in determining whether a person is depressed or whether the pain is real or psychological.

All these new features have sparked heated discussions regarding the invasion (or not) of privacy, since, for example, the identification of sexual preferences reached 81% of correctness by the algorithm, while people (without the resource) 61% of the time they made that identification correct.

There would therefore be a potential risk of discrimination on the part of companies, for example, in the recruitment of employees, so much so that legislators in some countries in Europe are already moving towards considering biometric data as information belonging to the people themselves and not something. in the public domain.

But, going back to the types of applications that would interest us, the fact is that facial expressions, most of the time, communicate more and better than words, either because they are spontaneous, sincere, natural, because they reflect the real feeling or emotion , and above all because they are not rational.

It is thus an important resource to be used in a complementary way to the other existing techniques in neuroscience and market research, in order to understand what people do not know, cannot or may not want to verbalize.

Paul Ekman was the pioneer in the analysis of facial expressions with his research even in the 60s. Among his greatest contributions, there is evidence of the existence of at least 6 human emotions, which can be identified regardless of gender, age or even of culture.

This study resulted in the decoding of these emotions in different combinations of 46 types of facial movements that allowed different applications from animations to lie detector.

With the advancement of technology and the greater capacity for processing information, the uses of this knowledge have expanded considerably. Basically the algorithm consists of measuring the movements of a series of points virtually created on the participant’s face in relation to other fixed points.

Thus, the position of the mouth, lips, cheeks, eyebrows, eyelids, wrinkles etc. they are constantly compared with other points (like the tip of the nose, the chin), with the pattern of the facial expression of each one and also with a reference table, allowing the identification of emotions such as joy, confusion, frustration, surprise, fear, anger, disgust, etc.

In studies that Checon has already carried out, it is possible to assess how much a particular commercial is pleasing (or not), to what extent this involvement is positive (or not), in which excerpts from the commercial are necessary adjustments or reinforcements, all without even asking the participants , always in a clear, objective way, which can be understood even by those who are not even in the area of ​​neuro or research.

The same types of conclusions are possible in the evaluation of people’s reactions to a political discourse, in order to understand if there is understanding or engagement, which are the passages that need more clarity, if there is approval, acceptance, always in a non-invasive way, without even not even interact directly with the participants.

In works on website usability, the difficulties in performing certain tasks are evident with negative facial expressions, or with expressions of anger, doubt, dissatisfaction or frustration, even if not admitted in the participants’ ‘rational’ testimony.

Another very interesting use of facial expressions is the analysis of consumers’ reactions to store windows, making it possible to identify the impact caused, the strengths and weaknesses of each option, even the attractiveness as an ‘invitation’ to people for the inside the establishment.

Not to mention the food industry analyzing facial expressions in tasting tests or preferably between products in blind tests, or in banks evaluating reactions to ATMs, or automakers evaluating the behavior of drivers driving or identifying signs of drowsiness behind the wheel.

Micro expressions offer a valuable contribution in the evaluation of decision making, being able to predict whether a certain purchase will be made (something that a flash of disgust can deny), or if a certain price is appropriate (something that an expression of joy can confirm and one of anger may disprove).

Since, in most cases, these micro expressions are not even perceived by the participants themselves, because they are spontaneous reactions, they are more reliable predictors than the (often) politically correct answers (not infrequently) that compromise many of the predictions of sales.

The choice of a candidate and the intention to vote also fit perfectly into this assessment of decision-making in a seemingly more reliable way than the mere rational statement by voters.

There is impressive data as to the correctness of the voting intentions of American voters in the 2010 presidential election, based only on the facial expressions collected in the Obama & Romney debate that reached 73% confirmation, without asking any questions.

More information:       The Economist – What machines can tell from your face

The New Yorker – We know how you feel

Paul Eckman – Facial Action Coding System