Skip to content
Variational Auto-encoder with Non-parametric Bayesian Prior
Python R
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
model
.gitignore
DPVAE.r
LICENSE
README.md
dp.py
gen_dp.png
gen_normal.png
report.markdown
vae_dp.py
vae_normal.py
vis1.png
vis2.png

README.md

VAE_NBP

Variational Auto-encoder with Non-parametric Bayesian Prior

Instruction:

   use vae_*.py to train the model.

For test and analysis you can use DPVAE.r. 

Examples:

source('DPVAE.r')

init('model')

z=sample(10)

image(display(z)[1,,])

image(display(z)[2,,])

image(display(z)[3,,])

image(display(z)[8,,])

z2=reconstruct(z)

image(display(z2)[8,,])

z2=reconstruct(z)

image(display(z2)[8,,])

image(display(z)[8,,])

image(display(z2)[8,,])

z2=reconstruct(z)

image(display(z2)[8,,])

z2=reconstruct(z)

image(display(z2)[8,,])

z2=reconstruct(z)

image(display(z2)[8,,])

z2=reconstruct(z)

image(display(z2)[8,,])

z2=reconstruct(z)

image(display(z2)[8,,])

z2=reconstruct(z)

image(display(z2)[8,,])
You can’t perform that action at this time.