论文标题
非线性ICA与一般不可压缩网络（GIN）的分解
Disentanglement by Nonlinear ICA with General Incompressible-flow Networks (GIN)
论文作者
论文摘要
代表学习的主要问题询问在哪个条件下可以重建任意复杂生成过程的真正潜在变量。 Khemakhem等人最近的突破性工作。 （2019年）在非线性ICA上为有条件的生成过程回答了这个问题。我们将这一重要结果扩展到与现实世界数据相关的方向。首先，我们将理论推广到未知的内在问题维度的情况下，并证明在某些特殊的（但不是非常限制的）情况下，信息的潜在变量将通过估计模型自动与噪声自动分离。此外，恢复的信息潜在变量将与生成过程的真实潜在变量一对一，直至琐碎的组件转换。其次，我们介绍了RealnVP可逆神经网络体系结构（Dinh等人（2016））的修改，该修改特别适合这种类型的问题：一般不可压缩流网络（GIN）。关于人工数据和Emnist的实验表明，在实践中确实验证了理论预测。特别是，我们提供了从Emnist提取的详细的22个信息的潜在变量。
A central question of representation learning asks under which conditions it is possible to reconstruct the true latent variables of an arbitrarily complex generative process. Recent breakthrough work by Khemakhem et al. (2019) on nonlinear ICA has answered this question for a broad class of conditional generative processes. We extend this important result in a direction relevant for application to real-world data. First, we generalize the theory to the case of unknown intrinsic problem dimension and prove that in some special (but not very restrictive) cases, informative latent variables will be automatically separated from noise by an estimating model. Furthermore, the recovered informative latent variables will be in one-to-one correspondence with the true latent variables of the generating process, up to a trivial component-wise transformation. Second, we introduce a modification of the RealNVP invertible neural network architecture (Dinh et al. (2016)) which is particularly suitable for this type of problem: the General Incompressible-flow Network (GIN). Experiments on artificial data and EMNIST demonstrate that theoretical predictions are indeed verified in practice. In particular, we provide a detailed set of exactly 22 informative latent variables extracted from EMNIST.
