BriefGPT.xyz
Jun, 2021
关于将归纳偏差引入VAE模型的研究
InteL-VAEs: Adding Inductive Biases to Variational Auto-Encoders via Intermediary Latents
HTML
PDF
Ning Miao, Emile Mathieu, N. Siddharth, Yee Whye Teh, Tom Rainforth
TL;DR
介绍了一种基于中介潜变量空间（InteL-VAEs）的有效方法，在生成式自编码器中直接引入归纳偏差特征，从而产生更好的生成模型和学习表征。
Abstract
We introduce a simple and effective method for learning
vaes
with controllable
inductive biases
by using an intermediary set of latent variables. This allows us to overcome the limitations of the standard Gaussia
→