A Demonstration of Human-Robot Communication Based on Multiskilled Language-Image Analysis

Abstract

Successful human-robot communication relies on strong machine-learning capabilities of natural language processing with contextual constraints. This paper developed and tested a communication platform using a social robot to investigate the applications of multiskilled language-image analysis. Three users tested the system to successfully complete a protocol based on referential communication tasks. The system may have broad applications in various scenarios involving human robot interaction.