Research Repository

Design of a Wearable Device for Reading Positive Expressions from Facial EMG Signals

Gruebler, Anna and Suzuki, Kenji (2014) 'Design of a Wearable Device for Reading Positive Expressions from Facial EMG Signals.' IEEE Transactions on Affective Computing, 5 (3). pp. 227-237. ISSN 1949-3045

Full text not available from this repository.


In this paper we present the design of a wearable device that reads positive facial expressions using physiological signals. We first analyze facial morphology in 3 dimensions and facial electromyographic signals on different facial locations and show that we can detect electromyographic signals with high amplitude on areas of low facial mobility on the side of the face, which are correlated to ones obtained from electrodes on traditional surface electromyographic capturing positions on top of facial muscles on the front of the face. We use a multi-attribute decision-making method to find adequate electrode positions on the side of face to capture these signals. Based on this analysis, we design and implement an ergonomic wearable device with high reliability. Because the signals are recorded distally, the proposed device uses independent component analysis and an artificial neural network to analyze them and achieve a high facial expression recognition rate on the side of the face. The recognized emotional facial expressions through the wearable interface device can be recorded during therapeutic interventions and for long-term facial expression recognition to quantify and infer the user's affective state in order to support medical professionals.

Item Type: Article
Uncontrolled Keywords: Electromyography; face and gesture recognition; pattern recognition; wearable interface
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: Faculty of Science and Health > Computer Science and Electronic Engineering, School of
SWORD Depositor: Elements
Depositing User: Elements
Date Deposited: 06 Jul 2015 13:32
Last Modified: 06 Jan 2022 14:39

Actions (login required)

View Item View Item