Home // International Journal On Advances in Software, volume 13, numbers 3 and 4, 2020 // View article

Facial Mimicry Analysis Based on 3D Morphable Face Models

Authors:
Oky Dicky Ardiansyah Prima
Yuta Ono
Hisayoshi Ito
Takahiro Tomisawa
Takashi Imabuchi

Keywords: mimicry; expression training; emotion; image processing.

Abstract:
Facial mimicry is an important non-verbal communication that can promote favorable social behavior and positive relationships. As recent computer vision technologies have reached the level of human perception for processing facial information, the study of automated analysis of facial mimicry has attracted the attention of the Human Computer Interaction (HCI) society. In this study, we propose a system to evaluate the similarity of facial images based on the shape of the face derived from 3-Dimensional (3D) face data. Two different 3D face data were used in this study: a 3D Digital Character (3DDC) and the Surrey 3D Morphable Face Model (3DMFM). Our approach consists of the following steps: (1) landmark extraction from the facial image; (2) 3D shape fitting; (3) similarity analysis for the face point cloud. Our results show that the similarity between faces can be assessed by analyzing the non-rigid portions of the faces. The proposed system can be extended as a facial mimicry training tool to improve social communication.

Pages: 274 to 283

Copyright: Copyright (c) to authors, 2020. Used with permission.

Publication date: December 30, 2020

Published in: journal

ISSN: 1942-2628