Latent Motion Manifold with Sequential Auto-Encoders

Simple item page
dc.contributor.authorJang, Deok-Kyeongen_US
dc.contributor.authorLee, Sung-Heeen_US
dc.contributor.editorSkouras, Melinaen_US
dc.date.accessioned2018-07-23T10:10:09Z
dc.date.available2018-07-23T10:10:09Z
dc.date.issued2018
dc.description.abstractWe propose the sequential autoencoders for constructing latent motion manifold. Sequential autoencoders minimize the difference between the ground truth motion space distribution and reconstructed motion space distribution sampled from the latent motion manifold. Our method is based on sequence-to-sequence model for encoding the temporal information of human motion. We also adopt Wasserstein regularizer for matching encoded training distribution to the prior distribution of motion manifold. Our experiments show that randomly sampled points from trained motion manifold distribution become natural and valid motions.en_US
dc.description.sectionheadersPosters
dc.description.seriesinformationEurographics/ ACM SIGGRAPH Symposium on Computer Animation - Posters
dc.identifier.doi10.2312/sca.20181184
dc.identifier.isbn978-3-03868-070-3
dc.identifier.issn1727-5288
dc.identifier.pages3-4
dc.identifier.urihttps://doi.org/10.2312/sca.20181184
dc.identifier.urihttps://diglib.eg.org:443/handle/10.2312/sca20181184
dc.publisherThe Eurographics Associationen_US
dc.subjectMathematics of computing
dc.subjectNonlinear equations
dc.titleLatent Motion Manifold with Sequential Auto-Encodersen_US
Files
Collections
SCA 18: Posters