对于深度生成模型,VAE和随机反向传播有什么区别?


Answers:


2

深度生成模型中的随机反向传播和近似推断似乎只是考虑了相同概念的更一般的实现。与VAE论文相比,他们:

  • 提出多组潜在变量,而不是只有一个的VAE

  • 考虑非对角的高斯后验,这使得优化更加困难

By using our site, you acknowledge that you have read and understand our Cookie Policy and Privacy Policy.
Licensed under cc by-sa 3.0 with attribution required.