Discriminator Configuration of Deep Learning Model Using Squared Connection

DOI

Bibliographic Information

Other Title
  • 2乗結合を用いた深層学習モデル識別部の構成

Abstract

Squared connection is a method that squares the transmission signals between neurons, enabling the confinement of the firing area of neurons to local regions in the feature space. By utilizing this method, the number of parameters and cimputational complexity can be reduced while maintaining its pattern recognition performance of deep learning models.

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top