How to implement & train a convolutional variational autoencoder with Keras in under 50 lines of code: https://colab.research.google.com/drive/1COeSoasrpY-w3uF1JPVksLEjx4Tx2ns5 …pic.twitter.com/MK580AGH53
Show this thread
1. The encoder & decoder 2. The VAE (with a custom train_step) 3. Training itpic.twitter.com/J8D4MjZWpt
-
-
Here's a slightly tweaked version that does a better job at showing the latent manifold of MNIST digits (I'll put this one in the next edition of my book to illustrate the manifold hypothesis)pic.twitter.com/3bx5LYwaNf
Show this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Literally came on twitter to take a break from building an autoencoder in Keras and this was at the top of my feed. What the heck man
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Thanks. I am currently developing a Denoising Convolutional AutoEncoder with Keras to build a Reverse Search Engine dedicated to soccer jerseys. Not yet there but making progress
pic.twitter.com/XgcLPgLSEO
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Thank you! One minor thing to note is that the `data` argument for train_step is a tuple, so `data[0]` should be used as a quick fix for dimension mismatching problem if you happen to have like me ; )
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.