mlsanityData to 7 Oct 2026

Circular-Symmetric Correlation Layer based on FFT

arXiv 2107.12480 ↗First posted 26 Jul 2021cs.CV cs.LGcited by 1

Not on the map. Why?

Abstract (arXiv)

Despite the vast success of standard planar convolutional neural networks, they are not the most efficient choice for analyzing signals that lie on an arbitrarily curved manifold, such as a cylinder. The problem arises when one performs a planar projection of these signals and inevitably causes them to be distorted or broken where there is valuable information. We propose a Circular-symmetric Correlation Layer (CCL) based on the formalism of roto-translation equivariant correlation on the continuous group S1 × ℝ, and implement it efficiently using the well-known Fast Fourier Transform (FFT) algorithm. We showcase the performance analysis of a general network equipped with CCL on various recognition and classification tasks and datasets. The PyTorch package implementation of CCL is provided online.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 22 of its 35 references that are in the map. First posted on arXiv 26 Jul 2021.

Builds on

22 of 35 references.

References · 22

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,880
  3. 2015Spatial Transformer Networkscited by 1,423
  4. 2017Deep Setscited by 1,394
  5. 2016Group Equivariant Convolutional Networkscited by 1,154
  6. 2014Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognitioncited by 719
  7. 2013Network In Networkcited by 600
  8. 2018Spherical CNNscited by 452
  9. 2018Invariant and Equivariant Graph Networkscited by 339
  10. 2016Steerable CNNscited by 298
  11. 2018On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groupscited by 296
  12. 2016Harmonic Networks: Deep Translation and Rotation Equivariancecited by 293
10 more
  1. 2014Understanding image representations by measuring their equivariance and equivalencecited by 240
  2. 2018A General Theory of Equivariant CNNs on Homogeneous Spacescited by 223
  3. 2017Equivariance Through Parameter-Sharingcited by 139
  4. 2020On Learning Sets of Symmetric Elementscited by 97
  5. 2016Exploiting Cyclic Symmetry in Convolutional Neural Networkscited by 91
  6. 2015Rotation-invariant convolutional neural networks for galaxy morphology predictioncited by 83
  7. 2020On the Universality of Rotation Equivariant Point Cloud Networkscited by 80
  8. 2020CyCNN: A Rotation Invariant CNN using Polar Mapping and Cylindrical Convolution Layerscited by 15
  9. 2016Permutation-equivariant neural networks applied to dynamics predictioncited by 14
  10. 2018SPNet: Deep 3D Object Classification and Retrieval using Stereographic Projectioncited by 8

Built on by

Cited by year

Citing papers per year: 2021: 1.

1 paper.

  1. 2021Equivariant Deep Dynamical Model for Motion Predictioncited by 2