alanaziyasir / vaim

VAIM

This is the implementation of Variational Autoencoder Inverse Mapper: An End-to-End Deep Learning Framework for Inverse Problems (VAIM). (https://ieeexplore.ieee.org/document/9534012)

Requirements

The code is written in Python=3.6, with the following libraries:

• tensorflow==1.11.0
• keras==2.1.2

Getting started

• Install the python libraries. (See Requirements).
• Download the code from GitHub:

git clone https://github.com/alanaziyasir/VAIM
cd VAIM

• Run the python script:

python3 train.py

• By default the script will run the first toy example which is f(x) = x^2.
• To run another example, adjust self.example variable in line 12 in VAIM.py.
• To see the jupyter notebbok demo go to VAIM_demo.ipynb.

Results:

• The script will create a directory saved_model/ and save the the weights with the lowest validation error
• It will also plot the latent and the results

Example plots after 5k epochs:

f(x) = x2	f(x) = sin(x)

latent of x2	latent of sin(x)