Neural Network Based Reinforcement Learning for Audio-Visual Gaze Control in Human-Robot Interaction

TL;DR

This paper presents a neural network-based reinforcement learning method enabling robots to autonomously learn and adapt gaze control strategies for human interaction using audio-visual cues without external sensors or human supervision.

Contribution

It introduces a novel combination of recurrent neural networks and Q-learning for autonomous, adaptive gaze control in robots, trained in simulation to avoid tedious real-world interactions.

Findings

Method is robust to parameter variations.

Joint audio-visual input yields best performance.

Experiments with Nao robot demonstrate social gaze behavior learning.

Abstract

This paper introduces a novel neural network-based reinforcement learning approach for robot gaze control. Our approach enables a robot to learn and to adapt its gaze control strategy for human-robot interaction neither with the use of external sensors nor with human supervision. The robot learns to focus its attention onto groups of people from its own audio-visual experiences, independently of the number of people, of their positions and of their physical appearances. In particular, we use a recurrent neural network architecture in combination with Q-learning to find an optimal action-selection policy; we pre-train the network using a simulated environment that mimics realistic scenarios that involve speaking/silent participants, thus avoiding the need of tedious sessions of a robot interacting with people. Our experimental evaluation suggests that the proposed method is robust against parameter estimation, i.e. the parameter values yielded by the method do not have a decisive impact on the performance. The best results are obtained when both audio and visual information is jointly used. Experiments with the Nao robot indicate that our framework is a step forward towards the autonomous learning of socially acceptable gaze behavior.