Show simple item record

dc.contributor.authorGrauman, Kristenen_US
dc.contributor.authorShakhnarovich, Gregoryen_US
dc.contributor.authorDarrell, Trevoren_US
dc.date.accessioned2004-10-08T20:38:51Z
dc.date.available2004-10-08T20:38:51Z
dc.date.issued2003-04-17en_US
dc.identifier.otherAIM-2003-008en_US
dc.identifier.urihttp://hdl.handle.net/1721.1/6714
dc.description.abstractWe present an image-based approach to infer 3D structure parameters using a probabilistic "shape+structure'' model. The 3D shape of a class of objects may be represented by sets of contours from silhouette views simultaneously observed from multiple calibrated cameras. Bayesian reconstructions of new shapes can then be estimated using a prior density constructed with a mixture model and probabilistic principal components analysis. We augment the shape model to incorporate structural features of interest; novel examples with missing structure parameters may then be reconstructed to obtain estimates of these parameters. Model matching and parameter inference are done entirely in the image domain and require no explicit 3D construction. Our shape model enables accurate estimation of structure despite segmentation errors or missing views in the input silhouettes, and works even with only a single input view. Using a dataset of thousands of pedestrian images generated from a synthetic model, we can perform accurate inference of the 3D locations of 19 joints on the body based on observed silhouette contours from real images.en_US
dc.format.extent17 p.en_US
dc.format.extent6362014 bytes
dc.format.extent9371703 bytes
dc.format.mimetypeapplication/postscript
dc.format.mimetypeapplication/pdf
dc.language.isoen_US
dc.relation.ispartofseriesAIM-2003-008en_US
dc.subjectAIen_US
dc.subject3D structureen_US
dc.subjectstatistical shape modelen_US
dc.subjectmulti-view imageryen_US
dc.subjectpose estimationen_US
dc.titleInferring 3D Structure with a Statistical Image-Based Shape Modelen_US


Files in this item

Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record