Memory-Friendly Deep Mesh Registration

Show simple item record
dc.contributor.authorLe Clerc, François
dc.contributor.authorSun, Hao
dc.contributor.editorSkala, Václav
dc.date.accessioned2020-07-27T07:54:30Z
dc.date.available2020-07-27T07:54:30Z
dc.date.issued2020
dc.description.abstract-translatedProcessing 3D meshes using convolutional neural networks requires convolutions to operate on features sampled on non-Euclidean manifolds. To this purpose, spatial-domain approaches applicable to meshes with different topologies locally map feature values in vertex neighborhoods to Euclidean ’patches’ that provide consistent inputs to the convolution filters around all mesh vertices. This generalization of the convolution operator significantly increases the memory footprint of convolutional layers and sets a practical limit to network depths on the available GPU hardware. We propose a memory-optimized convolution scheme that mitigates the issue and allows more convolutional layers to be included in a network for a given memory budget. The experimental evaluation of mesh registration accuracy on datasets of human face and body scans shows that deeper networks bring substantial performance improvements and demonstrate the benefits of our scheme. Our results outperform the state of art.en
dc.format10 s.cs
dc.format.mimetypeapplication/pdf
dc.identifier.citationWSCG 2020: full papers proceedings: 28th International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision, p. 1-10.en
dc.identifier.doihttps://doi.org/10.24132/CSRN.2020.3001.1
dc.identifier.isbn978-80-86943-35-0
dc.identifier.issn2464–4617 (print)
dc.identifier.issn2464–4625 (CD-ROM)
dc.identifier.urihttp://wscg.zcu.cz/WSCG2020/2020-CSRN-3001.pdf
dc.identifier.urihttp://hdl.handle.net/11025/38445
dc.language.isoenen
dc.publisherVáclav Skala - UNION Agencycs
dc.relation.ispartofseriesWSCG 2020: full papers proceedingsen
dc.rights© Václav Skala - UNION Agencycs
dc.rights.accessopenAccessen
dc.subjectgeometrické hluboké učenícs
dc.subjectkonvoluční neuronové sítěcs
dc.subjectpřizpůsobení tvarucs
dc.subject3D mřížkacs
dc.subject.translatedgeometric deep learningen
dc.subject.translatedconvolutional neural networksen
dc.subject.translatedshape matchingen
dc.subject.translated3D meshen
dc.titleMemory-Friendly Deep Mesh Registrationen
dc.typeconferenceObjecten
dc.typekonferenční příspěvekcs
dc.type.statusPeer-revieweden
dc.type.versionpublishedVersionen

Files

Original bundle
Showing 1 - 1 out of 1 results
No Thumbnail Available
Name:
E29.pdf
Size:
5.45 MB
Format:
Adobe Portable Document Format
Description:
Plný text
Download
License bundle
Showing 1 - 1 out of 1 results
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

WSCG 2020: Full Papers Proceedings