Abstract
In many real-word scenarios, humans and robots are required to coordinate their movements in joint tasks to fulfil a common goal. While several examples regarding dyadic human robot interaction exist in the current literature, multi-agent scenarios in which one or more artificial agents need to interact with many humans are still seldom investigated. In this paper we address the problem of synthesizing an autonomous artificial agent to perform a paradigmatic oscillatory joint task in human ensembles while exhibiting some desired human kinematic features. We propose an architecture based on deep reinforcement learning which is flexible enough to make the artificial agent interact with human groups of different sizes. As a paradigmatic coordination task we consider a multi-agent version of the mirror game, an oscillatory motor task largely used in the literature to study human motor coordination.
Highlights
The number of scenarios involving humans performing joint tasks with artificial agents is expected to grow rapidly in the near future
While different studies exist in the current literature involving dyadic coordination tasks between one human and one robot or avatar (Lamb et al, 2017; Peternel et al, 2017; Zhai et al, 2017), the problem of developing control-based cognitive architectures to drive autonomous artificial agents to interact with a human team remains much less investigated
We addressed the problem of synthesising an autonomous artificial agent able to coordinate its movement and perform a joint motor task in a group scenario
Summary
The number of scenarios involving humans performing joint tasks with artificial agents is expected to grow rapidly in the near future. Proposed in the seminal paper by Noy et al (2011), the mirror game in its original formulation involves two people coordinating the motion of their arm or finger so as to produce synchronous patterns This task has been largely used in the literature on interpersonal motor coordination and used to develop novel biomarkers for social disorders such as schizophrenia (Slowinski et al, 2014; Zhai et al, 2016; Zhai et al, 2017) but mostly in a dyadic coordination setting. The main drawback of this approach is the assumption that the other players in the group adjust their motion on a real time average of the positions of their neighbours This is clearly not the case with human players who tend to adjust their motion reciprocally in a number of different ways. To make the approach scalable, we present a training strategy which is independent from the number of players the CP is connected to while playing
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
