Autonomous detection of personality traits in human-robot interaction

• Grant number: 18/25876-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: June 01, 2019
• End date: December 31, 2019
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computer Systems
• Principal Investigator: Roseli Aparecida Francelin Romero
• Grantee: Victor Henrique de Souza Rodrigues
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

Most Human-Robot Interaction (HRI) studies incorporate social techniques based on human-human interaction in order to make our communication with robots more intuitive and efficient, especially in verbal interactions, which is the main means of human communication. Several works showed improvement in the interaction when the robot performs some adaptation or change of behavior throughout the interaction. Works with autonomous adaptation by means of audio are more common for the English language, marginalizing users who do not have domain of the same one. Thus, based on existing solutions for English language, this proposal aims to investigate the impact of an autonomous user profile detection system in an interactive architecture for Portuguese language, based on the audio collected in the interaction. A database with audios labeled in appropriate communication profiles will be created and deep neural networks will be used for autonomous detection of these profiles. Tests will be performed to validate the base and the detection algorithm, and also to integrate this method into an existing interactive robotic architecture.