Facial Emotion recognition analysis using deep learning through RGB-D imagery of VR participants through partially occluded facial types

PubDate: April 2022

Teams: Walton Institute for Information and Communication Systems Science

Writers: Ian Mills; Frances Cleary

PDF: Facial Emotion recognition analysis using deep learning through RGB-D imagery of VR participants through partially occluded facial types

Abstract

This research poster outlines the initial development of a facial emotion recognition (FER) evaluation system based on RGB-D imagery captured via a mobile based device. The study outlined features control group of non-occluded facial types and a set of participants wearing a head mounted display (HMD) in order to demonstrate an occluded facial type. We explore an architecture to develop a FER system that is suitable for occluded facial analysis. This paper details the methodology, experimental design and future work to be carried out to deliver such a system.