Projective Feature Learning for 3D Shapes with Multi-View Depth Images

Simple item page
dc.contributor.authorXie, Zhigeen_US
dc.contributor.authorXu, Kaien_US
dc.contributor.authorShan, Wenen_US
dc.contributor.authorLiu, Ligangen_US
dc.contributor.authorXiong, Yueshanen_US
dc.contributor.authorHuang, Huien_US
dc.contributor.editorStam, Jos and Mitra, Niloy J. and Xu, Kunen_US
dc.date.accessioned2015-10-07T05:11:54Z
dc.date.available2015-10-07T05:11:54Z
dc.date.issued2015en_US
dc.description.abstractFeature learning for 3D shapes is challenging due to the lack of natural paramterization for 3D surface models. We adopt the multi-view depth image representation and propose Multi-View Deep Extreme Learning Machine (MVD-ELM) to achieve fast and quality projective feature learning for 3D shapes. In contrast to existing multiview learning approaches, our method ensures the feature maps learned for different views are mutually dependent via shared weights and in each layer, their unprojections together form a valid 3D reconstruction of the input 3D shape through using normalized convolution kernels. These lead to a more accurate 3D feature learning as shown by the encouraging results in several applications. Moreover, the 3D reconstruction property enables clear visualization of the learned features, which further demonstrates the meaningfulness of our feature learning.en_US
dc.description.number7en_US
dc.description.sectionheadersShape and Meshen_US
dc.description.seriesinformationComputer Graphics Forumen_US
dc.description.volume34en_US
dc.identifier.doi10.1111/cgf.12740en_US
dc.identifier.pages001-011en_US
dc.identifier.urihttps://doi.org/10.1111/cgf.12740en_US
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.subjectI.3.5 [Computer Graphics]en_US
dc.subjectModelingen_US
dc.subjectComputational Geometry and Object Modelingen_US
dc.titleProjective Feature Learning for 3D Shapes with Multi-View Depth Imagesen_US
Files
Collections
34-Issue 7