BourGAN: Generative Networks with Metric Embeddings
BourGAN: Generative Networks with Metric Embeddings
复制标题
DOI：
--
复制
发表时间：
2018-05
期刊：
影响因子：
--
通讯作者：
Chang Xiao;Peilin Zhong;Changxi Zheng
中科院分区：
文献类型：
--
作者：
Chang Xiao;Peilin Zhong;Changxi Zheng
本文讨论了生成对抗网络（GAN）的模式崩溃。我们把模式看作是度量空间中数据分布的几何结构。在这种几何透镜下，我们将数据集的子样本从任意度量空间嵌入到l2空间中，同时保持它们的成对距离分布。这种度量嵌入不仅自动确定了潜在空间的维度，还使我们能够构建高斯混合来绘制潜在空间随机向量。我们使用高斯混合模型与目标函数的简单增强相结合来训练GAN。我们的方法的每个主要步骤都得到了理论分析的支持，我们对真实的和合成数据的实验证实，生成器能够产生分布在大多数模式上的样本，同时避免不需要的样本，在许多指标上优于最近的几个GAN变体，并提供新的功能。
This paper addresses the mode collapse for generative adversarial networks (GANs). We view modes as a geometric structure of data distribution in a metric space. Under this geometric lens, we embed subsamples of the dataset from an arbitrary metric space into the l2 space, while preserving their pairwise distance distribution. Not only does this metric embedding determine the dimensionality of the latent space automatically, it also enables us to construct a mixture of Gaussians to draw latent space random vectors. We use the Gaussian mixture model in tandem with a simple augmentation of the objective function to train GANs. Every major step of our method is supported by theoretical analysis, and our experiments on real and synthetic data confirm that the generator is able to produce samples spreading over most of the modes while avoiding unwanted samples, outperforming several recent GAN variants on a number of metrics and offering new features.