A Data-Efficient Deep Learning Approach for Deployable Multimodal Social Robots
This work addresses the challenge of deploying data-efficient deep learning for interactive social robots, though it is incremental with a focus on a specific game case study.
The authors tackled the problem of efficiently training a humanoid robot to play multimodal games like Noughts & Crosses, achieving high winning rates that substantially outperform DQN-based baselines with minimal data requirements of a few hundred images and demonstrations.
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 Pepper robot confirms that highly accurate visual perception is required for successful game play.