dpVAEs: Fixing Sample Generation for Regularized VAEs.
dpVAEs: Fixing Sample Generation for Regularized VAEs.
复制标题
DOI：
10.1007/978-3-030-69538-5_39
复制
发表时间：
2020-11
期刊：
影响因子：
--
通讯作者：
Elhabian S
中科院分区：
文献类型：
--
作者：
Bhalodia R;Lee I;Elhabian S
在没有标签数据的情况下，通过产生式建模的无监督表示学习是许多计算机视觉应用的主要内容。变分自动编码器(VAE)是一种强大的生成性模型，它学习对数据生成有用的表示法。然而，由于培训目标的内在挑战，虚拟企业无法学习适用于下游任务的有用表示法。基于正则化的方法试图改善VAE的表示学习方面，但代价是：糟糕的样本生成。在本文中，我们对正则化VAE的表示-生成权衡进行了探索，并引入了一种新的先验族，即解耦先验，即dpVAE，它将表示空间与生成空间解耦。这种解耦使得能够在表示空间上使用VAE正则化而不影响用于样本生成的分布，从而在不牺牲样本生成的情况下获得正则化的表示学习益处。DpVAE利用可逆网络来学习从任意复杂的表示分布到简单、易处理、可生成的分布的双射映射。解耦的先验可以适应最先进的VAE正则化，而不需要额外的超参数调整。我们展示了具有不同正则化的dpVAE的使用。在MNIST、SVHN和CelebA上的实验定量和定性地证明了dpVAE固定了规则化VAE的样本生成。
Unsupervised representation learning via generative modeling is a staple to many computer vision applications in the absence of labeled data. Variational Autoencoders (VAEs) are powerful generative models that learn representations useful for data generation. However, due to inherent challenges in the training objective, VAEs fail to learn useful representations amenable for downstream tasks. Regularization-based methods that attempt to improve the representation learning aspect of VAEs come at a price: poor sample generation. In this paper, we explore this representation-generation trade-off for regularized VAEs and introduce a new family of priors, namely decoupled priors, or dpVAEs, that decouple the representation space from the generation space. This decoupling enables the use of VAE regularizers on the representation space without impacting the distribution used for sample generation, and thereby reaping the representation learning benefits of the regularizations without sacrificing the sample generation. dpVAE leverages invertible networks to learn a bijective mapping from an arbitrarily complex representation distribution to a simple, tractable, generative distribution. Decoupled priors can be adapted to the state-of-the-art VAE regularizers without additional hyperparameter tuning. We showcase the use of dpVAEs with different regularizers. Experiments on MNIST, SVHN, and CelebA demonstrate, quantitatively and qualitatively, that dpVAE fixes sample generation for regularized VAEs.