Motion learning from observation using Affinity Propagation clustering

Guoting Chang, Dana Kulic

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

1 Citation (Scopus)


During robot imitation learning, a key problem when observing the motions of a demonstrator is the modeling and recognition of movement prototypes. This paper proposes using Affinity Propagation (AP) to cluster motions modeled using either Dynamic Movement Primitives (DMPs) or Hidden Markov Models (HMMs). The proposed AP clustering algorithm is simple and efficient, provides robust results and automatically identifies representative exemplars for each motion group, leading to a minimal representation of the observations that can also be used to generate motions. In experiments using videos and motion capture data of human demonstrations, it is shown that the weight parameters of the DMP model can be used as features for motion recognition and the proposed method can distinguish between different (coarse distinction) or similar (fine distinction) motion groups.

Original languageEnglish
Title of host publication22nd IEEE International Symposium on Robot and Human Interactive Communication
Subtitle of host publication"Living Together, Enjoying Together, and Working Together with Robots!", IEEE RO-MAN 2013
PublisherIEEE, Institute of Electrical and Electronics Engineers
Number of pages6
ISBN (Print)9781479905072
Publication statusPublished - 11 Dec 2013
Externally publishedYes
EventIEEE/RSJ International Symposium on Robot and Human Interactive Communication 2013 - Gyeongju, Korea, Republic of (South)
Duration: 26 Aug 201329 Aug 2013
Conference number: 22nd (Proceedings)

Publication series

NameProceedings - IEEE International Workshop on Robot and Human Interactive Communication


ConferenceIEEE/RSJ International Symposium on Robot and Human Interactive Communication 2013
Abbreviated titleRO-MAN 2013
CountryKorea, Republic of (South)
Internet address

Cite this