Crowdsourcing HRI through Online Multiplayer Games
Lists
Only you see this
Abstract
The development of hand-crafted action and dialog generation models for a social robot is a time consuming process that yields a solution only for the relatively narrow range of interactions envisioned by the programmers. In this paper, we propose a data-driven solution for interactive behavior generation that leverages online games as a means of collecting large-scale data corpora for human-robot interaction research. We present a system in which action and dialog models for a collaborative human-robot task are learned based on a reproduction of the task in a two-player online game called Mars Escape.
BibTeX
@inproceedings{chernova2010crowdsourcing,
author = {Sonia Chernova and Jeff Orkin and Cynthia Breazeal},
title = {{Crowdsourcing HRI through Online Multiplayer Games}},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2010},
url = {https://openreview.net/forum?id=d-deW-Lvmkq},
}Loading related content...