The Application that you can easily try it, submission for DL4US 3rd final task. Created to obtain a basis of Variational Auto Encoder and do an experiment in a reasoning ability of VAE. Enter a number from 1 ot 100 that shows an image and draw colored or random-noised boxes on left image. Then VAE reconstruct the edited image.

Run the following:

git clone https://github.com/yusuke1526/cat-faces-vae.git
cd cat-faces-vae
pip install -r requirements.txt
python main.py

also made slides about this.
https://drive.google.com/file/d/1g66wX5RG9_FVCXedN_nfbQg7YZiJ6GSz/view?usp=sharing