Abstract
We study two approaches for predicting an appropriate pose for a robot to take part in group formations typical of social human conversations subject to the physical layout of the surrounding environment. One method is model-based and explicitly encodes key geometric aspects of conversational formations. The other method is data-driven. It implicitly models key properties of spatial arrangements using graph neural networks and an adversarial training regimen. We evaluate the proposed approaches through quantitative metrics designed for this problem domain and via a human experiment. Our results suggest that the proposed methods are effective at reasoning about the environment layout and conversational group formations. They can also be used repeatedly to simulate conversational spatial arrangements despite being designed to output a single pose at a time. However, the methods showed different strengths. For example, the geometric approach was more successful at avoiding poses generated in nonfree areas of the environment, but the data-driven method was better at capturing the variability of conversational spatial formations. We discuss ways to address open challenges for the pose generation problem and other interesting avenues for future work.
Highlights
In this work, we study how to generate appropriate poses for social robots to take part in conversational group formations with users
We explore two approaches for generating spatial behavior: a model-based, geometric approach that explicitly encodes important properties of conversational group formations as often discussed in the social psychology literature (Kendon, 1990), and a data-driven adversarial approach that, once trained, implicitly encodes these properties
We analyzed the results for the quantitative metrics using restricted maximum likelihood (REML) analyses considering method (10 levels, each one corresponding to a row of Table 1) as main effect and Example ID from the Cocktail Party test set as random effect
Summary
We study how to generate appropriate poses for social robots to take part in conversational group formations with users. It is common to model conversational spatial behavior with discriminative models of group formations (Truong and Ngo, 2017; Vázquez et al, 2017; Hedayati et al, 2019; Barua et al, 2020; Swofford et al, 2020), we approach the problem of predicting a pose for a robot in a group conversation with generative models. Before explaining how the proposed methods work, the sections provide a brief introduction to conversational group formations from a social psychology perspective and introduce Graph Neural Networks (GNNs) from a message-passing point of view The former description is important for contextualizing the proposed geometric approach for pose generation and for understanding the rationale behind several of the metrics used in our evaluation. The formations keep groups as separate units from other close interactions
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
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 callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
