Home // ACHI 2020, The Thirteenth International Conference on Advances in Computer-Human Interactions // View article

Facial Mimicry Training Based on 3D Morphable Face Models

Authors:
Oky Dicky Ardiansyah Prima
Hisayoshi Ito
Takahiro Tomizawa
Takashi Imabuchi

Keywords: expression training; emotion; image processing

Abstract:
The recent techniques of automated facial expression recognition from facial images have achieved human perception levels. The application of this technology is expected not to be limited to facial expression analysis, but also to evaluate how well someone mimics another person’s expression. Facial mimic training will help people improve their interpersonal communication and that, in turn, will improve their work performance. This study proposes a self-learning-based expression training system using a simple 3D Morphable Face Model (3DMM). The proposed system analyzes faces of a subject and a given picture of a person who the subject is mimicking. The 68 facial landmarks for both faces are detected automatically and are used to fit a 3DMM using a deformation transfer technique. Our experiment shows that the proposed system accurately measures the similarity of facial appearance between subjects and their corresponding mimic targets. Thus, the proposed system can be used as a facial mimicry training tool to improve social communication.

Pages: 7 to 10

Copyright: Copyright (c) IARIA, 2020

Publication date: March 22, 2020

Published in: conference

ISSN: 2308-4138

ISBN: 978-1-61208-761-0

Location: Valencia, Spain

Dates: from November 21, 2020 to November 25, 2020