Weakly Supervised Disentanglement by Pairwise Similarities
Weakly Supervised Disentanglement by Pairwise Similarities
复制标题
DOI：
10.1609/aaai.v34i04.5754
复制
发表时间：
2019-06
期刊：
影响因子：
--
通讯作者：
Junxiang Chen;K. Batmanghelich
中科院分区：
文献类型：
--
作者：
Junxiang Chen;K. Batmanghelich
最近，与深度生成模型的无监督解纠缠学习相关的研究获得了广泛的关注。然而，在不引入监管的情况下，无法保证利益因素能够成功恢复（Locatello et al. 2018）。受现实世界问题的启发，我们提出了一种设置，用户通过基于要解开的因素提供实例之间的相似性来引入弱监督。相似性以二进制（是/否）或实值标签的形式提供，描述一对实例是否相似。我们提出了一种在变分自编码器框架内弱监督解开潜在变量的新方法。实验结果表明，利用弱监督可以显着提高解缠结方法的性能。
Recently, researches related to unsupervised disentanglement learning with deep generative models have gained substantial popularity. However, without introducing supervision, there is no guarantee that the factors of interest can be successfully recovered (Locatello et al. 2018). Motivated by a real-world problem, we propose a setting where the user introduces weak supervision by providing similarities between instances based on a factor to be disentangled. The similarity is provided as either a binary (yes/no) or real-valued label describing whether a pair of instances are similar or not. We propose a new method for weakly supervised disentanglement of latent variables within the framework of Variational Autoencoder. Experimental results demonstrate that utilizing weak supervision improves the performance of the disentanglement method substantially.