Empirische Inferenz Conference Paper 2012

Probabilistic Modeling of Human Movements for Intention Inference

Inference of human intention may be an essential step towards understanding human actions [21] and is hence important for realizing efficient human-robot interaction. In this paper, we propose the Intention-Driven Dynamics Model (IDDM), a latent variable model for inferring unknown human intentions. We train the model based on observed human behaviors/actions and we introduce an approximate inference algorithm to efficiently infer the human’s intention from an ongoing action. We verify the feasibility of the IDDM in two scenarios, i.e., target inference in robot table tennis and action recognition for interactive humanoid robots. In both tasks, the IDDM achieves substantial improvements over state-of-the-art regression and classification.

Author(s): Wang, Z. and Deisenroth, M. and Ben Amor, H. and Vogt, D. and Schölkopf, B. and Peters, J.
Book Title: Proceedings of Robotics: Science and Systems VIII
Pages: 8
Year: 2012
Day: 0
Bibtex Type: Conference Paper (inproceedings)
Event Name: R:SS 2012
Event Place: Sydney, Australia
State: Published
URL: http://www.roboticsproceedings.org/rss08/p55.html
Electronic Archiving: grant_archive
Links:

BibTex

@inproceedings{WangDBVSP2012,
  title = {Probabilistic Modeling of Human Movements for Intention Inference},
  booktitle = {Proceedings of Robotics: Science and Systems VIII},
  abstract = {Inference of human intention may be an essential step towards understanding human actions [21] and is hence
  important for realizing efficient human-robot interaction. In this paper, we propose the Intention-Driven Dynamics Model (IDDM), a latent variable model for inferring unknown human intentions. We train the model based on observed human behaviors/actions and we introduce an approximate inference algorithm to efficiently infer the human’s intention from an ongoing action.
  We verify the feasibility of the IDDM in two scenarios, i.e., target inference in robot table tennis and action recognition for interactive humanoid robots. In both tasks, the IDDM achieves substantial improvements over state-of-the-art regression and classification.},
  pages = {8},
  year = {2012},
  slug = {wangdbvsp2012},
  author = {Wang, Z. and Deisenroth, M. and Ben Amor, H. and Vogt, D. and Sch{\"o}lkopf, B. and Peters, J.},
  url = {http://www.roboticsproceedings.org/rss08/p55.html}
}