Home

  People

  Projects

  Software

  Publications

  Bachelor/Master

  Teaching

  Internal

  Links

  Contact

The Embodied Agents Research Group (EMBOTS) is a junior research group in the cluster of excellence Multimodal Computing and Interaction (MMCI) established by the German Research Foundation (DFG) within the framework of the German Excellence Initiative. The EMBOTS group is part of the German Research Center for AI (DFKI), Saarbrücken, Germany.

Mission

Our work aims at building multimodal human-computer interfaces, focussing on interactive embodied agents. One major challenge is to create the realistic animation of nonverbal behavior like gesture, gaze and posture shifts. To this end, we develop analysis tools, modeling methods and perform empirical validation of our results. The second challenge is to control the high-level behavior of embodied agents, managing user interaction and multimodal input/output. To this end, we design authoring tools that rely on a statechart-like visual interface and will ultimately be combined with plan- and rule-based approaches. In the past, members of our group have contributed both to foundational research and industrial applications. We have the ambition to bridge the gap between theoretical research and actual application, in order to empirically evaluate the end result.

Our group has co-organized the 9th International Conference on Intelligent Virtual Agents in Amsterdam (IVA 2009) and the 6th International Workshop on Multimodal Corpora on Malta (MMC 2010). In 2011, we initiated the International Workshop on Sign Language Translation and Avatar Technology in Berlin (SLTAT 2011).

Currently, we are co-organizing two events: the Multimodal Corpora Workshop at ICMI 2011 (submission deadline: 22 Aug 2011) and the 2nd International Workshop on Sign Language Translation and Avatar Technology (submission deadline: 17 July 2011).

Please visit our project page for details on our work and our software page for downloadable tools produced from our group.

News

Two more accepted full papers! Our submission Assessing the Deaf User Perspective on Sign Language Avatars has been accepted to ACM ASSETS (acceptance 30%), and the submission Using Individual Light Rigs to Control the Perception of Virtual Characters' Personality has been accepted to ACII (acceptance 34.6%).

Our submission Sign Language Avatars: Animation and Comprehensibility has been accepted as full paper at the 11th International Conference on Intelligent Virtual Agents (Reykjavik, Icelend). Acceptance rate: 19.6%.

The international workshop series on multimodal corpora, usually held in conjunction with the LREC conference, will be a sattelite event on the 13th International Conference on Multimodal Interaction (ICMI 2011) this year. Please visit our workshop website for the call and important dates.