We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Quantitative Biology > Quantitative Methods

Title: 6DCNN with roto-translational convolution filters for volumetric data processing

Abstract: In this work, we introduce 6D Convolutional Neural Network (6DCNN) designed to tackle the problem of detecting relative positions and orientations of local patterns when processing three-dimensional volumetric data. 6DCNN also includes SE(3)-equivariant message-passing and nonlinear activation operations constructed in the Fourier space. Working in the Fourier space allows significantly reducing the computational complexity of our operations. We demonstrate the properties of the 6D convolution and its efficiency in the recognition of spatial patterns. We also assess the 6DCNN model on several datasets from the recent CASP protein structure prediction challenges. Here, 6DCNN improves over the baseline architecture and also outperforms the state of the art.
Subjects: Quantitative Methods (q-bio.QM); Machine Learning (cs.LG); Signal Processing (eess.SP)
Cite as: arXiv:2107.12078 [q-bio.QM]
  (or arXiv:2107.12078v2 [q-bio.QM] for this version)

Submission history

From: Sergei Grudinin [view email]
[v1] Mon, 26 Jul 2021 09:56:55 GMT (1480kb,D)
[v2] Fri, 30 Jul 2021 08:04:05 GMT (1481kb,D)

Link back to: arXiv, form interface, contact.