3D steerable CNNs: Learning rotationally equivariant features in volumetric data

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

We present a convolutional network that is equivariant to rigid body motions. The model uses scalar-, vector-, and tensor fields over 3D Euclidean space to represent data, and equivariant convolutions to map between such representations. These SE(3)-equivariant convolutions utilize kernels which are parameterized as a linear combination of a complete steerable kernel basis, which is derived analytically in this paper. We prove that equivariant convolutions are the most general equivariant linear maps between fields over R^3. Our experimental results confirm the effectiveness of 3D Steerable CNNs for the problem of amino acid propensity prediction and protein structure classification, both of which have inherent SE(3) symmetry.
Original languageEnglish
Title of host publicationProceedings of the 32nd International Conference on Neural Information Processing Systems
Number of pages12
Volume2018
PublisherNIPS Proceedings
Publication date2018
Editionderc
Pages10381-10392
Publication statusPublished - 2018
Event32nd Annual Conference on Neural Information Processing Systems - Montreal, Montreal, Canada
Duration: 2 Dec 20188 Dec 2018
Conference number: 32
https://nips.cc/Conferences/2018

Conference

Conference32nd Annual Conference on Neural Information Processing Systems
Nummer32
LocationMontreal
LandCanada
ByMontreal
Periode02/12/201808/12/2018
Internetadresse
SeriesAdvances in Neural Information Processing Systems
Volume31
ISSN1049-5258

ID: 236511653