论文标题
ICE-BEEM：基于非线性ICA的可识别条件基于能量的深层模型
ICE-BeeM: Identifiable Conditional Energy-Based Deep Models Based on Nonlinear ICA
论文作者
论文摘要
我们考虑了概率模型的可识别性理论，并建立了足够的条件，在这些条件下，基于条件能量的模型非常广泛的代表在功能空间中是独一无二的，直到简单的转换。在我们的模型家族中，能量函数是两个特征提取器之间的点产物，一个用于因变量，一个用于调节变量。我们表明，在轻度条件下，这些特征是独特的缩放和排列。我们的结果扩展了非线性ICA的最新发展，实际上，它们导致了ICA模型的重要概括。特别是，我们表明我们的模型可用于估计独立调制组件分析框架（IMCA）中的组件，这是一种非线性ICA的新概括，可放宽独立性假设。一项彻底的实证研究表明，我们的模型从现实世界图像数据集中学到的表示形式是可识别的，并提高了转移学习和半监督学习任务的性能。
We consider the identifiability theory of probabilistic models and establish sufficient conditions under which the representations learned by a very broad family of conditional energy-based models are unique in function space, up to a simple transformation. In our model family, the energy function is the dot-product between two feature extractors, one for the dependent variable, and one for the conditioning variable. We show that under mild conditions, the features are unique up to scaling and permutation. Our results extend recent developments in nonlinear ICA, and in fact, they lead to an important generalization of ICA models. In particular, we show that our model can be used for the estimation of the components in the framework of Independently Modulated Component Analysis (IMCA), a new generalization of nonlinear ICA that relaxes the independence assumption. A thorough empirical study shows that representations learned by our model from real-world image datasets are identifiable, and improve performance in transfer learning and semi-supervised learning tasks.
