Open AccessSOGN: novel generative model using SOM
Author(s) -
Kim HoJoong,
Jung Sung Hoon
Publication year - 2019
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
ISSN - 1350-911X
DOI - 10.1049/el.2019.0202
Subject(s) - autoencoder , mnist database , generative grammar , artificial intelligence , generative model , computer science , pattern recognition (psychology) , artificial neural network , mode (computer interface) , convolutional neural network , image (mathematics) , generative adversarial network , operating system
Generative models such as variational autoencoder (VAE) and generative adversarial network (GAN) have been widely applied to many areas including image synthesis and voice generation. However, they have some problems that VAE makes blur images and GAN is difficult to learn due to mode collapsing. A novel generative model is proposed using a self‐organising map (SOM) termed a self‐organising generative network (SOGN). In the SOGN, training images are first mapped to SOM and then the output space of SOM is transformed into 2D vector spaces. These vector values are used as latent vectors to train the generative network such as artificial neural networks or convolutional neural networks. Experimental results with MNIST and CIFAR‐10 datasets showed that their generative model was easy to train without mode collapsing and made more clean images than VAE. It was also confirmed that the manifold is well‐observed without generating by the average effect of multiple images.