Troika GAN vs Decoupled GAN: An Investigation on the Impact of Subnetwork Weight Sharing for Data Augmentation

Notable advancements in the field of computer vision have transpired through the application of Generative Adversarial Networks (GANs). A new GAN variant, the Troika GAN (T-GAN), was recently proposed for data augmentation and was shown to be superior to the Coupled GAN (CoGAN) and the classic techn...

Full description

Saved in:
Bibliographic Details
Main Authors: Milan, Joe Anthony M, Fernandez, Proceso L, Jr
Format: text
Published: Archīum Ateneo 2020
Subjects:
Online Access:https://archium.ateneo.edu/discs-faculty-pubs/277
https://ieeexplore.ieee.org/document/9182445
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Ateneo De Manila University
id ph-ateneo-arc.discs-faculty-pubs-1278
record_format eprints
spelling ph-ateneo-arc.discs-faculty-pubs-12782022-03-18T07:12:10Z Troika GAN vs Decoupled GAN: An Investigation on the Impact of Subnetwork Weight Sharing for Data Augmentation Milan, Joe Anthony M Fernandez, Proceso L, Jr Notable advancements in the field of computer vision have transpired through the application of Generative Adversarial Networks (GANs). A new GAN variant, the Troika GAN (T-GAN), was recently proposed for data augmentation and was shown to be superior to the Coupled GAN (CoGAN) and the classic techniques of rotation and affine transformation. This paper describes our further investigation on T-GAN, specifically the impact of its subnetworks weight sharing. We decoupled the weight-sharing subnetworks of T-GAN to form three independent GANs, which we refer to collectively as the Decoupled GAN and whose weights are trained separately. We then used T-GAN and the Decoupled GAN to augment a set of words with limited instances from the IAM Handwriting Database. The resulting augmented datasets were applied to train the three types of Artificial Neural Network (ANN) classifiers: Vanilla ANN, Deep ANN, and Convolutional Neural Network (CNN). Results showed that the best accuracies from each of the 3 classifier types were obtained when these were trained with datasets augmented by a T-GAN. For example, the CNN classifier registered 89.76% as its best performance using T-GAN while recording only 87.47% accuracy from utilizing Decoupled GAN. A paired t-test between the 10-fold cross-validation results of these yielded a statistically significant p-value of 0.0075 in favor of the T-GAN augmentation. This clearly indicates that the sharing of weights is a vital factor in the generation of better synthetic data. With its significant impact on improving handwriting classification networks, T-GAN can be an ideal data augmentation approach to build robust systems where there is a scarcity of training dataset instances. 2020-01-01T08:00:00Z text https://archium.ateneo.edu/discs-faculty-pubs/277 https://ieeexplore.ieee.org/document/9182445 Department of Information Systems & Computer Science Faculty Publications Archīum Ateneo Gallium nitride Generative adversarial networks Artificial neural networks Training Generators Neurons Computer Sciences Databases and Information Systems
institution Ateneo De Manila University
building Ateneo De Manila University Library
continent Asia
country Philippines
Philippines
content_provider Ateneo De Manila University Library
collection archium.Ateneo Institutional Repository
topic Gallium nitride
Generative adversarial networks
Artificial neural networks
Training
Generators
Neurons
Computer Sciences
Databases and Information Systems
spellingShingle Gallium nitride
Generative adversarial networks
Artificial neural networks
Training
Generators
Neurons
Computer Sciences
Databases and Information Systems
Milan, Joe Anthony M
Fernandez, Proceso L, Jr
Troika GAN vs Decoupled GAN: An Investigation on the Impact of Subnetwork Weight Sharing for Data Augmentation
description Notable advancements in the field of computer vision have transpired through the application of Generative Adversarial Networks (GANs). A new GAN variant, the Troika GAN (T-GAN), was recently proposed for data augmentation and was shown to be superior to the Coupled GAN (CoGAN) and the classic techniques of rotation and affine transformation. This paper describes our further investigation on T-GAN, specifically the impact of its subnetworks weight sharing. We decoupled the weight-sharing subnetworks of T-GAN to form three independent GANs, which we refer to collectively as the Decoupled GAN and whose weights are trained separately. We then used T-GAN and the Decoupled GAN to augment a set of words with limited instances from the IAM Handwriting Database. The resulting augmented datasets were applied to train the three types of Artificial Neural Network (ANN) classifiers: Vanilla ANN, Deep ANN, and Convolutional Neural Network (CNN). Results showed that the best accuracies from each of the 3 classifier types were obtained when these were trained with datasets augmented by a T-GAN. For example, the CNN classifier registered 89.76% as its best performance using T-GAN while recording only 87.47% accuracy from utilizing Decoupled GAN. A paired t-test between the 10-fold cross-validation results of these yielded a statistically significant p-value of 0.0075 in favor of the T-GAN augmentation. This clearly indicates that the sharing of weights is a vital factor in the generation of better synthetic data. With its significant impact on improving handwriting classification networks, T-GAN can be an ideal data augmentation approach to build robust systems where there is a scarcity of training dataset instances.
format text
author Milan, Joe Anthony M
Fernandez, Proceso L, Jr
author_facet Milan, Joe Anthony M
Fernandez, Proceso L, Jr
author_sort Milan, Joe Anthony M
title Troika GAN vs Decoupled GAN: An Investigation on the Impact of Subnetwork Weight Sharing for Data Augmentation
title_short Troika GAN vs Decoupled GAN: An Investigation on the Impact of Subnetwork Weight Sharing for Data Augmentation
title_full Troika GAN vs Decoupled GAN: An Investigation on the Impact of Subnetwork Weight Sharing for Data Augmentation
title_fullStr Troika GAN vs Decoupled GAN: An Investigation on the Impact of Subnetwork Weight Sharing for Data Augmentation
title_full_unstemmed Troika GAN vs Decoupled GAN: An Investigation on the Impact of Subnetwork Weight Sharing for Data Augmentation
title_sort troika gan vs decoupled gan: an investigation on the impact of subnetwork weight sharing for data augmentation
publisher Archīum Ateneo
publishDate 2020
url https://archium.ateneo.edu/discs-faculty-pubs/277
https://ieeexplore.ieee.org/document/9182445
_version_ 1728621290891247616