This a TensorFlow Implementation of the Model Inversion Attack introduced with Model Inversion Attacks that Exploit Confidence Information and Basic Countermeasures (Fredrikson Et al.)
The gradient step and the final output of the attack loop is pre-processed with ZCA whitening and Global Contrast Normalization with Pylearn2, this helps to preserve the facial features present in the input dataset.
The important dependencies of this project include:
- TensorFlow
- Pylearn2
- Matplotlib
In case you run into some trouble installing the dependencies take a look at this issue.
- Download the AT&T Face Dataset from here
- Extract the dataset and replace the path variable in the 3rd cell of the inversion notebook.