Affective Computing

In the last few decades, we have seen rapid growth in technology that has helped us build smarter than ever.  Machines perform mechanical tasks ranging all over the spectrum from recognizing images to driving vehicles. Although movies and cartoons have been predicting the dawn of tech for years (like Baymax in Big Hero 6 capable of understanding human emotions) it’s only been a few years since trends have shifted toward actually developing machines that have a certain level of emotional awareness.

“Affective Computing” a term coined in the late 90’s by one of MIT’s computer scientists, Rosalind Picard; is a field of study at the intersection of Computer Science, Cognitive Science and Psychology that focuses on developing machines which can interpret and potentially simulate human ‘affects’. Her team’s latest work,  makes use of slight variations in facial expressions to determine the subject’s emotional state. The team was able to train a model which can be adapted to an entirely new group of people with minimal extra data and similar efficacy.

Meanwhile, another group of researchers was able to accurately predict personality traits for 42 students just by observing the movement of their eyes. The five major traits: ‘Extraversion’, ’Agreeableness’, “Openness’, ‘Conscientiousness’ and ‘Neuroticism’ in human psychology are used to describe how a person will react or make decisions in a situation. The team tracked their binocular gaze using a video based eye tracker to collect data, processed it using machine learning algorithms and compared results with a personality questionnaire taken by each student.

Studies like these demonstrate the direction we’re progressing toward and hint what future applications of affective computing could look like. A doctor may be able to monitor patients suffering from dementia, depression or someone on the autism spectrum easily. Or imagine your phone telling you to take a break because your eyes seem dry and tired. Perhaps after coming home from work one day, the refrigerator offers ice cream after detecting exhaustion … or more likely, takes it away to prevent stress eating!

Recently, Affectiva, a startup company that stemmed from Dr. Picard’s lab, released its first automotive application. The company is working with industry giants,  Porsche and BMW, to develop an application that will help cars detect if a driver is drowsy or distracted, and respond accordingly. This is just one of many possible applications of affective computing. However,  with more and more work being done in this field, there is an ongoing debate on the ethical limits of affective computing. Albeit, there are those who believe in the nobler side of this pursuit, we still need to accept that this technology will eventually be an advertiser’s paradise.