2026-09-18T17:25:43Zhttps://dspace.mit.edu/server/oai/requestoai:dspace.mit.edu:1721.1/1445682022-08-30T03:01:20Zcom_1721.1_7582com_1721.1_7581col_1721.1_131022
Solomon, Justin
Smirnov, Dmitriy
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
2022-08-29T15:56:25Z
2022-08-29T15:56:25Z
2022-05
2022-06-21T19:15:51.631Z
https://hdl.handle.net/1721.1/144568
0000-0002-6508-0705
While deep learning has been successfully applied to many tasks in computer graphics and vision, standard learning architectures often operate on shape representations that are dense and regular, like pixel or voxel grids. On the other hand, decades of computer graphics and geometry processing research have resulted in specialized algorithms and tools that use representations without such regular structure. In this thesis, we revisit conventional approaches in graphics in geometry to propose deep learning pipelines and inductive biases that are directly compatible with common geometry representations, without relying on simple uniform structure.
Ph.D.
Massachusetts Institute of Technology
In Copyright - Educational Use Permitted
Copyright MIT
http://rightsstatements.org/page/InC-EDU/1.0/
Deep Learning on Geometry Representations
Thesis
application/pdf
Doctoral
Doctor of Philosophy
Publication
unknown
unknown
<Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="f9ecb8ec-c719-45a8-907d-a1a49f41a5dc">
<Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type>
<Title>Deep Learning on Geometry Representations</Title>
<PublishedIn>
<Publication>
</Publication>
</PublishedIn>
<PublicationDate>2022-05</PublicationDate>
<Authors>
<Author>
<DisplayName>Smirnov, Dmitriy</DisplayName>
<Affiliation>
<OrgUnit>
</OrgUnit>
</Affiliation>
</Author>
</Authors>
<Editors>
</Editors>
<Publishers>
<Publisher>
<DisplayName>Massachusetts Institute of Technology</DisplayName>
<OrgUnit />
</Publisher>
</Publishers>
<License>http://rightsstatements.org/page/InC-EDU/1.0/</License>
<Abstract>While deep learning has been successfully applied to many tasks in computer graphics and vision, standard learning architectures often operate on shape representations that are dense and regular, like pixel or voxel grids. On the other hand, decades of computer graphics and geometry processing research have resulted in specialized algorithms and tools that use representations without such regular structure. In this thesis, we revisit conventional approaches in graphics in geometry to propose deep learning pipelines and inductive biases that are directly compatible with common geometry representations, without relying on simple uniform structure.</Abstract>
<Access xmlns="http://purl.org/coar/access_right"
>
</Access>
</Publication>