Scientists disagree on a definition of emotion, despite over a hundred years of research on the topic, but today most scientists agree that emotion has two main dimensions: arousal, ranging from calm to excited, and valence, ranging from negative to positive. Anger, for example, is likely to be highly arousing and negative in valence, while delight is likely to be mildly arousing and very positive in valence. While there is no perfect measure of arousal or valence, good approximations can be obtained by measuring arousal from the sympathetic nervous system and valence from facial expressions.
Dr. Rosalind Picard demonstrates two technologies for measuring emotional response that were invented at the MIT Media Lab and are being developed into products at Affectiva.
The sympathetic nervous system (SNS) is one of the three branches of the autonomic nervous system (ANS), along with the parasympathetic nervous system (PNS) and the enteric nervous system. The SNS is best known as the “fight or flight” system, but it has many additional roles. For example, SNS responds to positive emotion or to anticipation of something exciting that is about to happen. While most organs of the body are innervated by both the SNS and the PNS (for example heart rate is affected by a mix of SNS and PNS), the skin is mainly innervated by the SNS, making it an ideal place to measure sympathetic arousal.
The Affectiva Q™Sensor measures sympathetic arousal by measuring skin conductance. Skin conductance is a measure of electrodermal activity (EDA), sometimes called Galvanic skin response. The Q creates a tiny current across the surface of the skin and measures subtle electrical changes arising from the activity of the sweat glands, which “fill up" with increased cognitive and emotional activation.
While basic EDA sensing techniques have been used for a century (i.e. in lie detectors), methods for measuring EDA typically require electrodes, wires and a lab setting. Affectiva has changed the space of possibilities by making a comfortable sensor the size of a wristwatch, freeing people to wear it during everyday activities outside a lab. Portability and easy on-board logging or wireless streaming of the data greatly expand the ways people can now learn about sympathetic arousal.
Automated facial expression recognition is a relatively new and quickly evolving technology with its roots in the field of human computer interaction. Methods for recognizing facial expression have typically been developed by training the system on participants displaying exaggerated expressions of prototypical emotions, which limited its capabilities and made it unlikely to work in real-world applications.
Use Affectiva’s facial expression platform, Affdex, to recognize anatomical facial muscle movements called action units as well as to characterize collections of those movements together with larger motions of the head such as nods or shakes. Bayesian machine learning processes are used to combine the facial and head movements in order to recognize positive and negative displays of emotion as well as complex states such as interest and confusion.
Although some people claim that their face-reading technologies “recognize emotions,” it’s important to note that the state of the technology is such that it recognizes outward expressions, which may or may not correspond to true feelings. If you want outward expression to correspond to a participant’s inward feeling, then it is important to make sure the participant is truly comfortable expressing their inward feelings outwardly. Affective believes in opt-in technology and part of our ethic is that people should not be forced to have their emotions read if they do not want them read. We aim to respect people’s feelings in all our work. There are many times that people DO want to communicate their feelings, are happy to show genuine facial expressions, and these expressions may correspond to true feelings. In such cases, facial expression recognition can provide new, objective measures and lead to important insights in customer experience and other applications.
Emotion measurement can benefit people in many areas, including:
Market research: advertising, customer experience, distance learning, market research, product design, testing and usability feedback, sales promotion
Clinical research: addiction, affect dysregulation, alcoholism, Alzheimer’s Disease, anesthesia, anxiety disorder, autism spectrum disorders (ASD), attention deficit hyperactivity disorder (ADHD), depression, dysthymia, dermatology studies (e.g. eczema), diabetes, drug trials (providing an objective biomarker), endocrine disorders that interact with arousal levels (e.g. premenstrual dysmorphic disorder, hypoglycemia, hypothyroidism), epilepsy, hot flashes, menopause studies, Lou Gehrig’s disease or amyotrophic lateral sclerosis (ALS), locked-in syndrome, oppositional defiant disorder/conduct disorder, pain management, Parkinson’s disease, phobias (and desensitization therapy), post-traumatic stress disorder (PTSD), psychiatric counseling, quadriplegia or paraplegia, schizophrenia, sensory processing disorders, sexual dysfunction, sleep disorders, insomnia, sociopathy, stroke/temporary paralysis
Other: artistic expression, biofeedback, call centers, engagement measures, online gaming, learning style customization, nonverbal communication for nonspeaking people, robotics research, self-awareness enhancement, usability, virtual reality research