New deep approach learning techniques open opportunities to predict the mood of a set of individuals in a common environment.
Mexico City, February 6 (RT) .- A new focus that appeals to radio signals and artificial intelligence for detect and decipher human emotions, has been developed by a group of researchers from the University of London.
The method, published in the magazine PLOS ONE, uses wireless signals to measure heart rate and respiration to determine how a person is feeling without the need to consider any visual cues, such as a facial expression.
Study participants were asked to watch a video chosen by the researchers to elicit one of four basic emotions: anger, sadness, joy, and pleasure. During this process, wireless signals emitted by each person were safely captured. By analyzing their reactions, visible in slight body movements, the researchers were also able to read “covert” information in their heart and respiratory rhythms.
New research using radio waves to pick up subtle changes in heart rhythm and then, using an advanced AI to, understand what those signals means.
“Deep learning framework for subject-independent emotion detection using wireless signals” https://t.co/tjsq5AABT3
— Brian Roemmele (@BrianRoemmele) February 5, 2021
Previous studies also used non-invasive and wireless methods to detect emotions, but they relied on classic machine learning approaches, which use an algorithm to identify and classify emotional states. Unlike those, in this latest study the researchers used deep learning techniques – such as data analysis by an artificial neural network – that allowed them to detect emotions with greater precision.
“With deep learning we have shown that we can measure emotions accurately and independently of the subject: see a collection of signals from diverse individuals, learn from that data and use it to predict their emotions at a later stage,” said PhD student Achintha Ihalage, one of the authors.
They are now considering the use of low-cost systems, such as wi-fi routers, to detect the emotions of large numbers of people summoned, for example, to an office or other work environment.
“This research opens up many opportunities for practical applications, particularly in areas such as human-robot interaction, as well as health and emotional well-being, which has been increasingly important during the current COVID-19 pandemic,” explained the head of the project, Professor Yang Hao.
– .
