Learning Joint Policies for Human-Robot Dialog and Co-Navigation | IEEE Conference Publication | IEEE Xplore