InfoGAN-CR and ModelCentrality: Self-supervised Model Training and Selection for Disentangling GANs
InfoGAN-CR and ModelCentrality: Self-supervised Model Training and Selection for Disentangling GANs
复制标题
DOI：
--
复制
发表时间：
2019-06
期刊：
影响因子：
--
通讯作者：
Zinan Lin;K. K. Thekumparampil-K.;G. Fanti;Sewoong Oh
中科院分区：
文献类型：
--
作者：
Zinan Lin;K. K. Thekumparampil-K.;G. Fanti;Sewoong Oh
解纠缠生成模型将潜在代码向量映射到目标空间，同时强制学习的潜在代码的子集是可解释的并且与目标分布的不同属性相关联。最近的进展主要是基于变分自动编码器（VAE）的方法，而训练解纠缠生成对抗网络（GAN）仍然具有挑战性。在这项工作中，我们表明，可以通过使用自我监督来缓解解开GAN所面临的主要挑战。我们做了两个主要贡献：首先，我们设计了一种新的方法来训练具有自我监督的解纠缠GAN。我们提出了对比正则化，这是一个自然的概念的启发解开：潜在的遍历。这比最先进的基于VAE和GAN的方法获得了更高的解纠缠分数。其次，我们提出了一个无监督的模型选择方案，称为ModelCentrality，它使用生成的合成样本来计算一组模型的medoid（多维泛化中位数）。目前超参数调整的常见做法需要使用地面实况样本，每个样本都标记有已知的完美解纠缠潜码。由于真实的数据集没有配备这样的标签，我们提出了一个无监督的模型选择方案，并表明它找到了一个接近最好的模型，无论是VAE还是GAN。将对比正则化与ModelCentrality相结合，我们显着提高了最先进的解纠缠分数，而无需访问监督数据。
Disentangled generative models map a latent code vector to a target space, while enforcing that a subset of the learned latent codes are interpretable and associated with distinct properties of the target distribution. Recent advances have been dominated by Variational AutoEncoder (VAE)-based methods, while training disentangled generative adversarial networks (GANs) remains challenging. In this work, we show that the dominant challenges facing disentangled GANs can be mitigated through the use of self-supervision. We make two main contributions: first, we design a novel approach for training disentangled GANs with self-supervision. We propose contrastive regularizer, which is inspired by a natural notion of disentanglement: latent traversal. This achieves higher disentanglement scores than state-of-the-art VAE- and GAN-based approaches. Second, we propose an unsupervised model selection scheme called ModelCentrality, which uses generated synthetic samples to compute the medoid (multi-dimensional generalization of median) of a collection of models. The current common practice of hyper-parameter tuning requires using ground-truths samples, each labelled with known perfect disentangled latent codes. As real datasets are not equipped with such labels, we propose an unsupervised model selection scheme and show that it finds a model close to the best one, for both VAEs and GANs. Combining contrastive regularization with ModelCentrality, we improve upon the state-of-the-art disentanglement scores significantly, without accessing the supervised data.