Gesture semantics reconstruction based on motion capturing and complex event processing

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Insa Lawler, Assistant Professor (Creator)
Institution
The University of North Carolina at Greensboro (UNCG )
Web Site: http://library.uncg.edu/

Abstract: A fundamental problem in manual based gesture semantics reconstruction is the specification of preferred semantic concepts for gesture trajectories. This issue is complicated by problems human raters have annotating fast-paced three dimensional trajectories. Based on a detailed example of a gesticulated circular trajectory, we present a data-driven approach that covers parts of the semantic reconstruction by making use of motion capturing (mocap) technology. In our FA3ME framework we use a complex event processing approach to analyse and annotate multi-modal events. This framework provides grounds for a detailed description of how to get at the semantic concept of circularity observed in the data.

Additional Information

Publication
14th Annual Meeting of the Special Interest Group on Discourse and Dialogue, 270-279.
Language: English
Date: 2013
Keywords
semantics, complex event processing, gesture annotation, circular trajectory

Email this document to