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

Toward Automated Analysis of Communication Mirroring

Authors:
Kumiko Hosogoe
Miyu Nakano
Oky Dicky Ardiansyah Prima
Yuta Ono

Keywords: communication mirroring; human pose estimation; DTW; 360 degree camera

Abstract:
Mirroring is a technique in which someone unconsciously reflects other people's behavior, such as gestures and facial expressions. This technique can help us to build rapport with others by making communication more effective and reflective. Due to the developments in computer vision, human behavior observation based on vision cameras has become viable. Moreover, the wide spread of omnidirectional cameras has made it possible to observe more people at the same time, making it easier to analyze face-to-face conversation scenes. In this study, we propose a framework to perform a time series analysis based on Dynamic Time Warping (DTW) to determine whether communication mirroring has been established. The framework uses human pose estimation techniques to track hand gestures of two persons during a conversation. The framework will detect all gestures that are similar to the trained gestures. Our experiments show that the DTW was able to detect mirroring acts having distinct gestures. However, detections of similarities of weak gestures are challenging.

Pages: 15 to 18

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