arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2010.03549cs.CVcs.AI
收藏

基于判别模型的生成对抗网络评价

On the Evaluation of Generative Adversarial Networks By Discriminative Models

Virginia Tech(弗吉尼亚理工大学)

Amirsina Torfi, Mohammadreza Beyki, Edward A. Fox

摘要:

生成对抗网络(GANs)可以准确地对复杂的多维数据建模并生成真实的样本。然而,由于它们对数据分布的隐式估计,它们的评估是一项具有挑战性的任务。大多数与解决这一问题相关的研究工作都是通过定性视觉评估进行验证的。这样的方法不能很好地推广到图像域之外。由于其中许多评估指标是针对视觉领域提出的,因此很难应用于其他领域。量化措施是必要的,以更好地指导不同GANs模型的训练和比较。在这项工作中,我们利用Siamese神经网络提出了一个领域不可知的评估指标:(1)具有与人类评估一致的定性评估,(2)相对于常见的GAN问题(如模式丢弃和发明)是鲁棒的,(3)不需要任何预先训练的分类器。本文的实证结果表明,该方法的优越性相比,流行的Inception评分和竞争力的FID评分。

英文摘要:

Generative Adversarial Networks (GANs) can accurately model complex multi-dimensional data and generate realistic samples. However, due to their implicit estimation of data distributions, their evaluation is a challenging task. The majority of research efforts associated with tackling this issue were validated by qualitative visual evaluation. Such approaches do not generalize well beyond the image domain. Since many of those evaluation metrics are proposed and bound to the vision domain, they are difficult to apply to other domains. Quantitative measures are necessary to better guide the training and comparison of different GANs models. In this work, we leverage Siamese neural networks to propose a domain-agnostic evaluation metric: (1) with a qualitative evaluation that is consistent with human evaluation, (2) that is robust relative to common GAN issues such as mode dropping and invention, and (3) does not require any pretrained classifier. The empirical results in this paper demonstrate the superiority of this method compared to the popular Inception Score and are competitive with the FID score.

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。

↑