Biometrics and the Body
When we are afraid the heart races, breathing becomes rapid, the mouth dries up, muscles tense, and palms become sweaty. We may feel anxious, stressed, panicked or nervous, but emotion recognition does not end here. Psychophysiology shows us that such changes are mediated by the autonomic nervous system (ANS), which operates the sympathetic and parasympathetic divisions of the human body.
The sympathetic nervous system (SNS) is a “quick response mobilizing system” responsible for “fight or flight” responses, which prepare the body to react to stresses such as threat or injury. It directs muscles to contract and heart rate to increase. The parasympathetic nervous system (PNS) is a “more slowly activated dampening system” referred to as “rest and digest”, which controls functions of the body at rest. It helps maintain homeostasis by directing muscles to relax and heart rate to decrease. In conjunction, the two constitute what psychophysiologists refer to as the ANS.
Although both the sympathetic and parasympathetic divisions drive opposing effects on the body, it is the balance of activity between the two that helps maintain an internal stable environment in the face of changing external stimuli and conditions. During an anxious experience for instance, the body will divert blood flow from parasympathetic nerve functions (such as digestion) to sympathetic nerve functions (such as muscle contraction and heavy breathing). There is little people can do to consciously control their PNS, but there are factors (such as exercise and experience) that can help some people exert a level of control over the sympathetic responses.
Certain emotional states affect this balance and can result in a wide variety of bodily reactions similar to the ones described in brackets above. Importantly, these bodily reactions can be monitored and measured through signals referred to as bio-signals. All we can observe from the outside are the bodily reactions. Bio-sensors give us insights into how and why these reactions are occurring across the interplay between parasympathetic and sympathetic nerve functions. As a result, detection and analysis of the these involuntary, subconscious divisions of the body are becoming an increasingly important field for Human Computer Interaction (HCI) as the advantages of emotional recognition and machine learning become more apparent and achievable online and across everyday life.
As they can be made to operate unobtrusively and robustly against a number of environmental and social conditions which other forms of emotion recognition have difficulty overcoming, a bio-signal focused approach offers important methodological advantages.
First, while more traditional qualitative emotion research uses interviews, questionnaires, and expert opinions, these focus largely on examining subjective feelings, and can be limited in explaining, and do not allow for real time measurements. In this sense, they tend to be utilized most effectively before or after an emotional state is experienced. Second, in contrast to Brain Computer Interfacing (BCI), which remains highly costly, immobile and obtrusive, biosignals are much less invasive, mobile, and can be recorded and processed in real time across different spaces. This makes biosignals rich, affordable and scalable sources of information for mental health and other emotion-related research.
Despite these advantages, the relationship between emotion and bio-signals remains an under-explored aspect of the myriad of scientific research being done using machine learning processes. According to an important report by Sony Corporate Laboratories Europe, this is because machine learning and human-computer interaction are traditionally viewed as logical and rational mechanisms, which are thought to be incompatible with the supposedly irrational nature of emotions.
Long past are the days when emotions were thought to be irrational experiences we can only hope to grasp cognitively after the fact. Since the advent of “Affective Computing” in the early 2000s, research into learning, information retrieval, communications, entertainment, design, mental health, and HCI have focused on the key role affects and emotions play in understanding phenomena such as attention, memory, and aesthetics. In particular, with wearable biosignals that perceive context and environment, as well as physiological information, there is the promise of gathering powerful data for improving our understanding of factors that contribute to human health and well-being.
To bring emotional awareness, wellness, and mental health care into the modern age in innovative and accessible ways, organizations can and should utilize the complimentary advantages bio-sensorial technologies offer, especially when coupled with qualitative emotion research methods and general self-reporting. As affordable, scalable, mobile apparatuses, bio-signals present a unique opportunity to examine how our built, natural, and social environments aggregate or buffer against mental health issues.
