A Data-Efficient Deep Learning Approach for Deployable Multimodal Social Robots

Cuayahuitl, Heriberto (2019) A Data-Efficient Deep Learning Approach for Deployable Multimodal Social Robots. Neurocomputing . ISSN 0925-2312

Documents
A Data-Efficient Deep Learning Approach for Deployable Multimodal Social Robots

Request a copy
[img] PDF
main.pdf - Whole Document
Restricted to Repository staff only until 16 September 2019.
Available under License Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International.

11MB
Item Type:Article
Item Status:Live Archive

Abstract

The deep supervised and reinforcement learning paradigms (among others) have the potential to endow interactive multimodal social robots with the ability of acquiring skills autonomously. But it is still not very clear yet how they can be best deployed in real world applications. As a step in this direction, we propose a deep learning-based approach for efficiently training a humanoid robot to play multimodal games---and use the game of `Noughts \& Crosses' with two variants as a case study. Its minimum requirements for learning to perceive and interact are based on a few hundred example images, a few example multimodal dialogues and physical demonstrations of robot manipulation, and automatic simulations. In addition, we propose novel algorithms for robust visual game tracking and for competitive policy learning with high winning rates, which substantially outperform DQN-based baselines. While an automatic evaluation shows evidence that the proposed approach can be easily extended to new games with competitive robot behaviours, a human evaluation with 130 humans playing with the {\it Pepper} robot confirms that highly accurate visual perception is required for successful game play.

Additional Information:The final published version of this article can be accessed online at https://www.journals.elsevier.com/neurocomputing/
Keywords:Deep Reinforcement Learning, Deep Supervised Learning, Interactive Robots, Multimodal Perception and Interaction, Board Games
Subjects:H Engineering > H671 Robotics
G Mathematical and Computer Sciences > G700 Artificial Intelligence
G Mathematical and Computer Sciences > G730 Neural Computing
Divisions:College of Science > School of Computer Science
ID Code:33533
Deposited On:11 Oct 2018 14:46

Repository Staff Only: item control page