arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.11432cs.LGstat.ML
收藏

基于在线核的生成对抗网络

Online Kernel based Generative Adversarial Networks

Georgia Institute of Technology(佐治亚理工学院);Carnegie Mellon University(卡内基梅隆大学)

Yeojoon Youn, Neil Thistlethwaite, Sang Keun Choe, Jacob Abernethy

摘要:

在过去五年中,深度学习的重大突破之一是生成对抗网络(GAN),这是一种基于神经网络的生成模型,旨在模拟给定样本数据集的一些底层分布。与许多有监督的问题不同,在这些问题中,人们试图最小化参数的简单目标函数,GAN训练被公式化为一对网络参数上的最小-最大问题。虽然经验上GANs在几个领域取得了令人印象深刻的成功,但研究人员一直对不寻常的训练行为感到困惑,包括循环所谓的模式坍塌。在本文中,我们首先提供了一种定量方法来探索GAN训练中的一些挑战,并以经验的方式展示了这与神经网络的参数性质之间的关系。我们提出了一种新的方法,解决了许多这些问题,依赖于一个基于内核的非参数神经网络,是非常适合在线培训-我们称之为在线基于内核的生成对抗网络(OKGAN)。我们的经验表明,OKGAN减轻了一些训练问题,包括模式坍塌和循环,并且更符合理论保证。在合成数据上,OKGAN在反向KL发散方面的经验表现明显优于其他GAN公式;在MNIST、SVHN和CelebA等经典视觉数据集上,表现出相当的性能。

英文摘要:

One of the major breakthroughs in deep learning over the past five years has been the Generative Adversarial Network (GAN), a neural network-based generative model which aims to mimic some underlying distribution given a dataset of samples. In contrast to many supervised problems, where one tries to minimize a simple objective function of the parameters, GAN training is formulated as a min-max problem over a pair of network parameters. While empirically GANs have shown impressive success in several domains, researchers have been puzzled by unusual training behavior, including cycling so-called mode collapse. In this paper, we begin by providing a quantitative method to explore some of the challenges in GAN training, and we show empirically how this relates fundamentally to the parametric nature of the discriminator network. We propose a novel approach that resolves many of these issues by relying on a kernel-based non-parametric discriminator that is highly amenable to online training---we call this the Online Kernel-based Generative Adversarial Networks (OKGAN). We show empirically that OKGANs mitigate a number of training issues, including mode collapse and cycling, and are much more amenable to theoretical guarantees. OKGANs empirically perform dramatically better, with respect to reverse KL-divergence, than other GAN formulations on synthetic data; on classical vision datasets such as MNIST, SVHN, and CelebA, show comparable performance.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑