Deep learning based social context perception on human-robot interaction
Journal
49th International Symposium on Robotics, ISR 2017
Date Issued
2017
Author(s)
Hsieh C.K.
Abstract
The objective of this paper is to develop a social co-robot for provision of "just-good services" using situational context based perception for perceiving human's mentation. In the near future, more and more human-friendly robots will be involved in human livelihood. In order to act more friendly in Human Social Environments (HSEs), robots should possess the ability of social context perception. In this paper, we propose a deep learning model, which learn from previous observations of human-robot interaction, to enable our robot to perceive a human's needs. Therefore, the appropriate social behaviors with respect to human's mental state would be reacted by robot. The experimental results demonstrate that the proposed deep learning model can significantly improve the accuracy of predicting a person's mentation from human behaviors. ? 49th International Symposium on Robotics, ISR 2017. All rights reserved.
Subjects
Behavioral research; Deep learning; Man machine systems; Robotics; Good services; Human behaviors; Human-Friendly Robots; Learning models; Situational context; Social behavior; Social context; Social environment; Human robot interaction
Type
conference paper