The Automatic Multimodal Robotic Storyteller

Abstract: This repository contains the full implementation of the Automatic Multimodal Robotic Storyteller for the Pepper robot. The system enables Pepper to perform any written story using multimodal storytelling elements such as emotional body language, emotion-inducing music, and dynamic colored lighting. It consists of two main modules: an annotator module that enriches story texts with emotional metadata, and a player module that translates these annotations into expressive robotic behavior. The pipeline has been empirically developed based on extensive research into emotional storytelling and robotic expressivity. It is designed to be easily adaptable to a wide variety of stories, allowing users to convert plain text into compelling robotic performances with minimal effort. This tool aims to support the growing interest in social robots as engaging storytelling agents and provides a practical foundation for future applications and research in the field, as internet connection is only needed for the annotator but not the player.