Geared Rotationally Identical and Invariant Convolutional Neural Network Systems

no code implementations • 3 Aug 2018 • ShihChung B. Lo, Matthew T. Freedman, M. D., Seong K. Mun, Ph. D., Heang-Ping Chan, Ph. D

Theorems and techniques to form different types of transformationally invariant processing and to produce the same output quantitatively based on either transformationally invariant operators or symmetric operations have recently been introduced by the authors.

Transformationally Identical and Invariant Convolutional Neural Networks by Combining Symmetric Operations or Input Vectors

no code implementations • 30 Jul 2018 • ShihChung B. Lo, Matthew T. Freedman, Seong K. Mun

Other transformationally identical CNNs can be constructed by averaging transformed input vectors of the family at the input layer followed by an ordinary CNN process or by a set of symmetric operations.