Skip to content
Lossy image autoencoders with convolution and deconvolution networks in Tensorflow
Branch: master
Clone or download
Pull request Compare This branch is even with giuseppebonaccorso:master.
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.
.gitignore
LICENSE
Lossy Image Autoencoder.ipynb
README.md

README.md

Lossy image autoencoders with convolution and deconvolution networks in Tensorflow


This Jupyter notebook refers to: https://www.bonaccorso.eu/2017/07/29/lossy-image-autoencoders-convolution-deconvolution-networks-tensorflow/

Requirements

  • Python 2.7-3.5
  • Tensorflow
  • Keras
  • SciPy
  • Scikit-Image
  • Numba (optional)

Example with CIFAR-10 dataset

(Trained with CIFAR-10 dataset (with 50000 samples) and a code length equal to 128)

First row: original images, second row: lossy reconstructions

Possible improvements

Possible improvements include:

  • Adding a flag (using a placeholder) to use the model for both training and prediction. In the former mode, the input is an image batch, while in the latter is a code batch
  • Using L1 (and/or L2) code regularization
You can’t perform that action at this time.