DO-GAN: A double oracle framework for generative adversarial networks

In this paper, we propose a new approach to train Gen-erative Adversarial Networks (GANs) where we deploy a double-oracle framework using the generator and discrim-inator oracles. GAN is essentially a two-player zero-sum game between the generator and the discriminator. Training GANs is challenging...

Full description

Saved in:
Bibliographic Details
Main Authors: AUNG, Aye Phyu Phye, WANG, Xinrun, YU, Runsheng, AN, Bo, JAYAVELU, Senthilnath, LI, Xiaoli
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2022
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/9136
https://ink.library.smu.edu.sg/context/sis_research/article/10139/viewcontent/DO_GAN_CVPR_2022_av.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
id sg-smu-ink.sis_research-10139
record_format dspace
spelling sg-smu-ink.sis_research-101392024-08-01T09:26:46Z DO-GAN: A double oracle framework for generative adversarial networks AUNG, Aye Phyu Phye WANG, Xinrun YU, Runsheng AN, Bo JAYAVELU, Senthilnath LI, Xiaoli In this paper, we propose a new approach to train Gen-erative Adversarial Networks (GANs) where we deploy a double-oracle framework using the generator and discrim-inator oracles. GAN is essentially a two-player zero-sum game between the generator and the discriminator. Training GANs is challenging as a pure Nash equilibrium may not exist and even finding the mixed Nash equilibrium is difficult as GANs have a large-scale strategy space. In DO-GAN, we extend the double oracle framework to GANs. We first generalize the players' strategies as the trained models of generator and discriminator from the best response or-acles. We then compute the meta-strategies using a linear program. For scalability of the framework where multi-ple generators and discriminator best responses are stored in the memory, we propose two solutions: 1) pruning the weakly-dominated players' strategies to keep the oracles from becoming intractable; 2) applying continual learning to retain the previous knowledge of the networks. We apply our framework to established GAN architectures such as vanilla GAN, Deep Convolutional GAN, Spectral Normalization GAN and Stacked GAN. Finally, we conduct experiments on MNIST, CIFAR-10 and CelebA datasets and show that DO-GAN variants have significant improvements in both subjective qualitative evaluation and quantitative metrics, compared with their respective GAN architectures. 2022-06-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/9136 info:doi/10.1109/CVPR52688.2022.01099 https://ink.library.smu.edu.sg/context/sis_research/article/10139/viewcontent/DO_GAN_CVPR_2022_av.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Deep learning architectures and techniques Image and video synthesis and generation Optimization methods Artificial Intelligence and Robotics Theory and Algorithms
institution Singapore Management University
building SMU Libraries
continent Asia
country Singapore
Singapore
content_provider SMU Libraries
collection InK@SMU
language English
topic Deep learning architectures and techniques
Image and video synthesis and generation
Optimization methods
Artificial Intelligence and Robotics
Theory and Algorithms
spellingShingle Deep learning architectures and techniques
Image and video synthesis and generation
Optimization methods
Artificial Intelligence and Robotics
Theory and Algorithms
AUNG, Aye Phyu Phye
WANG, Xinrun
YU, Runsheng
AN, Bo
JAYAVELU, Senthilnath
LI, Xiaoli
DO-GAN: A double oracle framework for generative adversarial networks
description In this paper, we propose a new approach to train Gen-erative Adversarial Networks (GANs) where we deploy a double-oracle framework using the generator and discrim-inator oracles. GAN is essentially a two-player zero-sum game between the generator and the discriminator. Training GANs is challenging as a pure Nash equilibrium may not exist and even finding the mixed Nash equilibrium is difficult as GANs have a large-scale strategy space. In DO-GAN, we extend the double oracle framework to GANs. We first generalize the players' strategies as the trained models of generator and discriminator from the best response or-acles. We then compute the meta-strategies using a linear program. For scalability of the framework where multi-ple generators and discriminator best responses are stored in the memory, we propose two solutions: 1) pruning the weakly-dominated players' strategies to keep the oracles from becoming intractable; 2) applying continual learning to retain the previous knowledge of the networks. We apply our framework to established GAN architectures such as vanilla GAN, Deep Convolutional GAN, Spectral Normalization GAN and Stacked GAN. Finally, we conduct experiments on MNIST, CIFAR-10 and CelebA datasets and show that DO-GAN variants have significant improvements in both subjective qualitative evaluation and quantitative metrics, compared with their respective GAN architectures.
format text
author AUNG, Aye Phyu Phye
WANG, Xinrun
YU, Runsheng
AN, Bo
JAYAVELU, Senthilnath
LI, Xiaoli
author_facet AUNG, Aye Phyu Phye
WANG, Xinrun
YU, Runsheng
AN, Bo
JAYAVELU, Senthilnath
LI, Xiaoli
author_sort AUNG, Aye Phyu Phye
title DO-GAN: A double oracle framework for generative adversarial networks
title_short DO-GAN: A double oracle framework for generative adversarial networks
title_full DO-GAN: A double oracle framework for generative adversarial networks
title_fullStr DO-GAN: A double oracle framework for generative adversarial networks
title_full_unstemmed DO-GAN: A double oracle framework for generative adversarial networks
title_sort do-gan: a double oracle framework for generative adversarial networks
publisher Institutional Knowledge at Singapore Management University
publishDate 2022
url https://ink.library.smu.edu.sg/sis_research/9136
https://ink.library.smu.edu.sg/context/sis_research/article/10139/viewcontent/DO_GAN_CVPR_2022_av.pdf
_version_ 1814047752417443840