Code to shredder-v1: Learning Noise for Privacy with Partial DNN Inference on the Edge by FatemehSadat Mireshghallah ([email protected])

The paper pdf is available at: https://arxiv.org/pdf/1905.11814v1.pdf, note that it is version1 of the work, not the final version.

This work appeared in the thirty-sixth International Conference on Machine Learning (ICML19), Security and Privacy of Machin Learning Workshop (https://icml2019workshop.github.io/).

In this repository you can find the code to shredder, and also the .npy files created through sampling, so you do not need to run everything from scratch, you can use those.

1. To do noise training, and save trained samples, run "alexnet-activation.py". This is a script generated from the more descriptive "alexnet-activation.ipynb" notebook. This generates a .npy file that has samples in it. Since this is a one time thing and takes a while, we have provided this named "activation-4-alexnet.npy"

2. To sample from the trained noise and save activations for calculating the mutual information, run sample-for-mutual-info-alexnet.py. The results of this step are also provided, with the names: noisy-activation-4-laplace-MI.npy, original-activation-4-laplace-MI.npy, and original-image-4-laplace-MI.npy which is over 100 mb (around 600mb) and we had to upload it to "https://ufile.io/jhz2d8r7"

3. To see the Mutual Info, you should first have the ITE toolbox cloned (https://bitbucket.org/szzoli/ite-in-python/src/default/). Then, run notebook "mutual_info_ITE-laplace-04.ipynb"

Please do not hesitate to contact me in case of any issues

This software is Copyright © 2019 The Regents of the University of California. All Rights Reserved. Permission to copy, modify, and distribute this software and its documentation for educational, research and non-profit purposes, without fee, and without a written agreement is hereby granted, provided that the above copyright notice, this paragraph and the following three paragraphs appear in all copies. Permission to make commercial use of this software may be obtained by contacting:

Office of Innovation and Commercialization

9500 Gilman Drive, Mail Code 0910

University of California

La Jolla, CA 92093-0910

(858) 534-5815

[email protected]

This software program and documentation are copyrighted by The Regents of the University of California. The software program and documentation are supplied “as is”, without any accompanying services from The Regents. The Regents does not warrant that the operation of the program will be uninterrupted or error-free. The end-user understands that the program was developed for research purposes and is advised not to rely exclusively on the program for any reason.

IN NO EVENT SHALL THE UNIVERSITY OF CALIFORNIA BE LIABLE TO ANY PARTY FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING LOST PROFITS, ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN IF THE UNIVERSITY OF CALIFORNIA HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. THE UNIVERSITY OF CALIFORNIA SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS ON AN “AS IS” BASIS, AND THE UNIVERSITY OF CALIFORNIA HAS NO OBLIGATIONS TO PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.