Abstract
Inference of human intention may be an essential step towards understanding human actions 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 movements/actions. We introduce an efficient approximate inference algorithm to infer the human's intention from an ongoing movement. 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.
| Original language | English (US) |
|---|---|
| Title of host publication | Robotics: Science and Systems |
| Publisher | MIT Press Journals |
| Pages | 433-440 |
| Number of pages | 8 |
| Volume | 8 |
| ISBN (Print) | 9780262519687 |
| State | Published - 2013 |
| Externally published | Yes |
| Event | International Conference on Robotics Science and Systems, RSS 2012 - Sydney, Australia Duration: Jul 9 2012 → Jul 13 2012 |
Other
| Other | International Conference on Robotics Science and Systems, RSS 2012 |
|---|---|
| Country | Australia |
| City | Sydney |
| Period | 7/9/12 → 7/13/12 |
Fingerprint
ASJC Scopus subject areas
- Artificial Intelligence
- Control and Systems Engineering
- Electrical and Electronic Engineering
Cite this
Probabilistic modeling of human movements for intention inference. / Wang, Zhikun; Deisenroth, Marc Peter; Ben Amor, Hani; Vogt, David; Schölkopf, Bernhard; Peters, Jan.
Robotics: Science and Systems. Vol. 8 MIT Press Journals, 2013. p. 433-440.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Probabilistic modeling of human movements for intention inference
AU - Wang, Zhikun
AU - Deisenroth, Marc Peter
AU - Ben Amor, Hani
AU - Vogt, David
AU - Schölkopf, Bernhard
AU - Peters, Jan
PY - 2013
Y1 - 2013
N2 - Inference of human intention may be an essential step towards understanding human actions 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 movements/actions. We introduce an efficient approximate inference algorithm to infer the human's intention from an ongoing movement. 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.
AB - Inference of human intention may be an essential step towards understanding human actions 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 movements/actions. We introduce an efficient approximate inference algorithm to infer the human's intention from an ongoing movement. 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.
UR - http://www.scopus.com/inward/record.url?scp=84959297900&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84959297900&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:84959297900
SN - 9780262519687
VL - 8
SP - 433
EP - 440
BT - Robotics: Science and Systems
PB - MIT Press Journals
ER -