Emote: A Wearable Device for Emotion

Estimation and Classification

Presenter: Jorge Romero

Co-Presenter(s):
Oscar, Avendano

Presenter Status: Undergraduate student

Academic Year: 22-23

Semester: Spring

Faculty Mentor: Mohamed Salem

Department: Engineering

Funding Source/Sponsor: Koret Scholars Program

President's Strategic Plan Goal: Connectivity and Community Engagement

Abstract:
The ability for automatic estimation and classification of human emotions can be crucial for certain individuals with disabilities or difficulties in communication, such as some of the non-verbal individuals on the autistic spectrum. Therefore, a wearable device for emotion detection and classification is essential in determining the individual's emotional state. To achieve the desired results based on our research, we integrated various biometric circuits [Blood Oxygen Levels, Heart Rate, Skin Conductivity Response, and Skin Temperature] in an attempt to correlate these changes to four basic human emotions: Fear, Happiness, Sadness, and Anger. By conducting a blind experiment on multiple participants and surveying their actual emotional responses, we are able to correlate the behavior of each data set based on the deviation away from the baseline readings. As a result, we will be able to synthesize a method that allows us to interpret the emotional state of the individual not only by being non-invasive but also by allowing behavior specialists or guardians to keep watch on the individual should they need assistance.