Communications of the ACM
Scientists Propose App That Detects Emotions Based on Walking Style
Researchers at the University of the Chinese Academy of Sciences are using portable electronic devices to read the emotional states of their users, based on the way they walk.
Credit: TheStack.com
Researchers at the University of the Chinese Academy of Sciences are developing a methodology for identifying a person's emotional state based on the way they walk, in hopes this could be incorporated into wearable devices and apps.
The researchers fitted 59 students with a pair of Samsung Galaxy S2 smartphones, which measured the movement of their wrists and ankles as they walked around a rectangular course for two minutes, after which they were asked to watch an emotionally priming video and then return to the track. The subjects reported notable changes to their emotional state after viewing the videos, and the researchers found these changes registered in the gait data being gathered by the smartphones. The movement of the ankle was found to be the most clearly tied to the wearer's emotional state, and the researchers found the greatest classification accuracy was yielded by the support vector machine-learning method model.
Although this methodology will have to be validated, the researchers say it could have a wide range of applications, including integration with health apps or personal-assistant apps such as Siri and Cortana.
From The Stack (UK)
View Full Article
Abstracts Copyright © 2015 Information Inc., Bethesda, Maryland, USA
No entries found