Show, Attend and Interact: Perceivable Human-Robot Social Interaction through Neural Attention Q-Network

TL;DR

This paper presents a neural attention-based reinforcement learning system enabling a robot to learn human-like social interaction skills through 14 days of real-world interactions, resulting in perceivable and socially acceptable responses.

Contribution

It introduces the Multimodal Deep Attention Recurrent Q-Network that learns social behaviors via end-to-end reinforcement learning from real-world human interactions.

Findings

Robot learned to respond to complex human behaviors.

System demonstrated perceivable and socially acceptable responses.

Achieved natural social interaction after 14 days.

Abstract

For a safe, natural and effective human-robot social interaction, it is essential to develop a system that allows a robot to demonstrate the perceivable responsive behaviors to complex human behaviors. We introduce the Multimodal Deep Attention Recurrent Q-Network using which the robot exhibits human-like social interaction skills after 14 days of interacting with people in an uncontrolled real world. Each and every day during the 14 days, the system gathered robot interaction experiences with people through a hit-and-trial method and then trained the MDARQN on these experiences using end-to-end reinforcement learning approach. The results of interaction based learning indicate that the robot has learned to respond to complex human behaviors in a perceivable and socially acceptable manner.

Full text

See pages 1-last of ICRA_2017_Cam.pdf