# Demonstration-Guided Deep Reinforcement Learning of Control Policies for Dexterous Human-Robot Interaction

## TL;DR

This paper introduces a deep reinforcement learning approach for training control policies enabling humanoid robots to perform natural human-like hand interactions, using a parameterizable reward function based on motion capture data.

## Contribution

It presents a novel parameterizable reward function for learning diverse human-robot hand interactions without changing the reward structure.

## Key findings

- Policies produce natural-looking motions
- Large-scale user study confirms human-like perception
- Method successfully applied to handshake, hand clap, finger touch

## Abstract

In this paper, we propose a method for training control policies for human-robot interactions such as handshakes or hand claps via Deep Reinforcement Learning. The policy controls a humanoid Shadow Dexterous Hand, attached to a robot arm. We propose a parameterizable multi-objective reward function that allows learning of a variety of interactions without changing the reward structure. The parameters of the reward function are estimated directly from motion capture data of human-human interactions in order to produce policies that are perceived as being natural and human-like by observers. We evaluate our method on three significantly different hand interactions: handshake, hand clap and finger touch. We provide detailed analysis of the proposed reward function and the resulting policies and conduct a large-scale user study, indicating that our policy produces natural looking motions.