Modern life is increasingly exposing human beings to stressful situations with can have severe health ramifications.
According to the Gallup 2019 Global Emotions Report 55% of American adults reported their behaviours have been negatively affected due to the physical and emotional toll of increased stress, while the global average of the number of stressed people out of 143 countries is 35%.
Stress is associated with a wide array of potential health problems ranging from anxiety and depression to digestive issues, with this emotional state also a factor in potentially fatal health issues such as heart attacks and strokes.
The elevation of stress levels across the globe has made access to stress management strategies more important than ever before. One of the key factors in managing stress is developing better ways of monitoring it.
One indicator of stress may be unconscious movements like teeth grinding, leg tapping, or hand rubbing. These subtle movements or expressions across the body that occur involuntarily and often unconsciously are called “micro gestures” and can be used to gauge emotional states, even in cases when subjects are attempting to hide these states.
In the past, scientists have investigated the connection between behaviours like deception and micro gestures, but the link between subconscious movements and emotional states, and stress in particular has been less well explored.
In a new paper published in the journal Advanced Intelligent Systems, the authors, including Leo-Le Fang, researcher at the Laboratory for Artificial Intelligence in Design, Hong Kong detail the use of an emerging technology for wearable systems called EmoSense to detect micro gestures that could be associated with stress.
“EmoSense is a novel three-layer stress detection system framework that leverages a capacitive sensing technique to capture micro gestures, even those that go unnoticed or are unconscious signals of stress,” said Fang. “It adopts a machine learning algorithm to track these micro gestures accurately.”
EmoSense has a three-layer structure with the top layer designed to adjust conductivity, the middle layer being fully conductive, and the bottom layer being fully resistive.
This helps to insulate the system against unwanted human contact as EmoSense uses the conductivity of the body and the resistive nature of the skin to measure how alternating current (AC) passing through the body, changes as a result of tissue changes triggered by micro gestures.
“Our study presents a pilot user experiment that demonstrates the effective tracking of micro gestures and their role in gauging stress levels,” Fang added. “We propose an efficient micro gesture detection system using capacitive sensing and machine learning algorithms.”
Under Pressure: Putting EmoSense to the Test
In the pilot study to assess a link between micro gestures and stress, the team tested EmoSense with 16 participants who were asked to report their emotional states during the experiments.
“Through user studies, participants’ micro gestures can be tracked while concurrently gathering self-reported stress levels,” Fang explained. “Then the quantitative statistical analysis can be conducted to uncover a significant positive correlation between micro gesture frequency and stress level.”
Based on this, the team was able to determine a mathematical model linking micro gestures to stress response. What they found was individuals experiencing high levels of stress or nervousness exhibited unconscious hand-based micro gestures, and these even occurred when the subjects were able to control their facial or vocal expressions consciously.
“Our user studies further revealed a significant discovery — a moderate correlation exists between stress levels and micro gesture frequency,” Fang explained, indicating that how often micro gestures occur could be an indicator of how stressed a person feels.
The user reports collected by the team from their subjects included other terms to describe negative emotional states such as “Fearful” and “Annoyed.” This observation has sparked the team’s curiosity in exploring the underlying correlations between these negative emotions and other micro gestures, Fang pointed out.
The team is also considering how EmoSense could be deployed in wearable electronics such as wristbands and watchbands, with the three-layer design of the system advantageous to preventing false signals created by mistouching.
“Moving forward, we plan to apply EmoSense to a range of practical applications, design prototypes, and conduct user studies to assess its effectiveness and usability. Additionally, building on the intriguing observation mentioned earlier, our future work will focus on exploring the hand-based responses of various emotions.”
Reference: Le Fang., et al., EmoSense: Revealing True Emotions Through Microgestures, Advanced Intelligent Systems, (2023), DOI: 10.1002/aisy.202300050