BriefGPT.xyz
Nov, 2016
利用 Householder 流改进变分自编码器
Improving Variational Auto-Encoders using Householder Flow
HTML
PDF
Jakub M. Tomczak, Max Welling
TL;DR
本论文提出了一种基于Householder变换的体积保存流方法,改进了变分后验分布的灵活性并在MNIST数据集和组织病理学数据中获得了理想的结果。
Abstract
variational auto-encoders
(VAE) are scalable and powerful generative models. However, the choice of the variational posterior determines tractability and flexibility of the VAE. Commonly,
latent variables
are mod
→