%0 Journal Article %T Multi-Mode Tensor Representation of Motion Data %A Krüger, Björn %A Tautges, Jochen %A Müller, Meinard %A Weber, Andreas %J JVRB - Journal of Virtual Reality and Broadcasting %D 2008 %V 5(2008) %N 5 %@ 1860-2037 %F krüger2008 %X In this paper, we investigate how a multilinear modelcan be used to represent human motion data. Basedon technical modes (referring to degrees of freedomand number of frames) and natural modes that typicallyappear in the context of a motion capture session(referring to actor, style, and repetition), the motiondata is encoded in form of a high-order tensor.This tensor is then reduced by using N-mode singularvalue decomposition. Our experiments show that thereduced model approximates the original motion betterthen previously introduced PCA-based approaches.Furthermore, we discuss how the tensor representationmay be used as a valuable tool for the synthesis of newmotions. %L 004 %K Motion Capture %K Motion Synthesis %K N-Mode SVD %K Tensor Representation %R 10.20385/1860-2037/5.2008.5 %U http://nbn-resolving.de/urn:nbn:de:0009-6-14197 %U http://dx.doi.org/10.20385/1860-2037/5.2008.5