Speech driven embodied interaction systems called "InterRobot" as physical media and "InterActor" as electronic media are developed for supporting human interaction and communicaion by generating the coherently related communicative motions to speech input. Talkers can share the context of communication by mutual embodied entrainment via "InterRobot" and "InterActor". In this paper, the effectiveness of the proposed hierarchical interaction model for "InterRobot" and "InterActor" is confirmed by comparing with single stage model. Actual applications of "InterRobot" and "InterActor" to human interface are also demonstrated for developing a new embodied communication industry.