A Novel Convolutional Neural Network Architecture with a Continuous Symmetry. (arXiv:2308.01621v2 [cs.CV] UPDATED)

This paper introduces a new Convolutional Neural Network (ConvNet)
architecture inspired by a class of partial differential equations (PDEs)
called quasi-linear hyperbolic systems. With comparable performance on the
image classification task, it allows for the modification of the weights via a
continuous group of symmetry. This is a significant shift from traditional
models where the architecture and weights are essentially fixed. We wish to
promote the (internal) symmetry as a new desirable property for a neural
network, and to draw …

architecture arxiv classification continuous convolutional neural network differential image linear network network architecture neural network novel paper performance symmetry systems