HalfedgeCNN for Native and Flexible Deep Learning on Triangle Meshes
DC Element | Wert | Sprache |
---|---|---|
dc.contributor.author | Ludwig, I. | |
dc.contributor.author | Tyson, D. | |
dc.contributor.author | Campen, M. | |
dc.date.accessioned | 2024-01-04T10:32:57Z | - |
dc.date.available | 2024-01-04T10:32:57Z | - |
dc.date.issued | 2023 | |
dc.identifier.issn | 0167-7055 | |
dc.identifier.uri | http://osnascholar.ub.uni-osnabrueck.de/handle/unios/73087 | - |
dc.description.abstract | We describe HalfedgeCNN, a collection of modules to build neural networks that operate on triangle meshes. Taking inspiration from the (edge-based) MeshCNN, convolution, pooling, and unpooling layers are consistently defined on the basis of halfedges of the mesh, pairs of oppositely oriented virtual instances of each edge. This provides benefits over alternative definitions on the basis of vertices, edges, or faces. Additional interface layers enable support for feature data associated with such mesh entities in input and output as well. Due to being defined natively on mesh entities and their neighborhoods, lossy resampling or interpolation techniques (to enable the application of operators adopted from image domains) do not need to be employed. The operators have various degrees of freedom that can be exploited to adapt to application-specific needs. | |
dc.description.sponsorship | Deutsche~Forschungsgemeinschaft (DFG) [456666331]; The authors wish to thank Steffen Hinderink for helpful discussions and the authors of MeshCNN for open-sourcing their implementation and data. This work was supported by the Deutsche & nbsp;Forschungsgemeinschaft (DFG) - 456666331. Open Access funding enabled and organized by Projekt DEAL. | |
dc.language.iso | en | |
dc.publisher | WILEY | |
dc.relation.ispartof | COMPUTER GRAPHICS FORUM | |
dc.subject | & RARR | |
dc.subject | CCS Concepts | |
dc.subject | Computer Science | |
dc.subject | Computer Science, Software Engineering | |
dc.subject | Computing methodologies | |
dc.subject | Mesh models | |
dc.subject | Neural networks | |
dc.subject | Shape analysis | |
dc.title | HalfedgeCNN for Native and Flexible Deep Learning on Triangle Meshes | |
dc.type | journal article | |
dc.identifier.doi | 10.1111/cgf.14898 | |
dc.identifier.isi | ISI:001046120900001 | |
dc.description.volume | 42 | |
dc.description.issue | 5 | |
dc.identifier.eissn | 1467-8659 | |
dc.publisher.place | 111 RIVER ST, HOBOKEN 07030-5774, NJ USA | |
dcterms.isPartOf.abbreviation | Comput. Graph. Forum | |
dcterms.oaStatus | hybrid | |
local.import.remains | affiliations : University Osnabruck | |
local.import.remains | earlyaccessdate : AUG 2023 | |
local.import.remains | web-of-science-index : Science Citation Index Expanded (SCI-EXPANDED) | |
crisitem.author.dept | FB 06 - Mathematik/Informatik | - |
crisitem.author.deptid | fb06 | - |
crisitem.author.orcid | 0000-0003-2340-3462 | - |
crisitem.author.parentorg | Universität Osnabrück | - |
crisitem.author.netid | CaMa281 | - |
Seitenaufrufe
2
Letzte Woche
0
0
Letzter Monat
0
0
geprüft am 29.05.2024