Jamie Hayes

Information Security Group - Computer Science Department - University College London

Jamie Hayes
Computer Science Department
University College London
Gower Street
London WC1E 6BT
United Kingdom
Email: j.hayes at cs.ucl.ac.uk
(GPG) 3AAD77570EF22F6C

Current Projects

  • Adversarial Machine Learning.
  • Further work on Website Fingerprinting.

I am a student studying for a PhD at the Information Security group at UCL. My supervisor is George Danezis and second supervisor is Thore Graepel.

My research interests include adversarial machine learning, network traffic analysis, privacy, anonymity and machine learning (ML) applications to information security. Before joining UCL I studied under Anuj Dawar in the Programming, Logic, and Semantics Group at the University of Cambridge, working on the parameterized complexity of the graph isomorphism problem.

Recent News

  • I spent the summer at Microsoft Research and Google Research, and our paper on multi-party machine learning was accepted to NIPS 2018.
  • My privacy-preserving machine learning pipeline has recently gone live on GOV.UK Verify! Check back here soon for the whitepaper.
  • I have been invited to give at talk at IBM Thomas J. Watson Research Center on adversarial machine learning.
  • I have been awarded a Google Fellowship in Machine Learning.
  • I am now an external reviewer for the Privacy Enhancing Technologies Symposium 2017. Please consider submitting your work.
  • Recently, I have been participating in root-me CTFs and war games. Root Me is a platform for everyone to test and improve knowledge in computer security and hacking, and I would definitely recommend checking it out. I have also been working my way through CryptoPals challenges. You can find my solutions here.
  • Download a (semi) useful extension.
  • Our paper on Website Fingerprinting Defenses has been accepted to PETS '17. You can find the paper, code and dataset on our Tor Hidden Service (HS). The HS is currently running the defense proposed within the paper.
  • See what computers think people look like here.
  • During the first quarter of 2017 I will be working at the GDS on device fingerprinting problems, and the NRL on traffic analysis.




  • On Visible Adversarial Perturbations & Digital Watermarking. CVPR Workshop Track 2018.
  • Learning Universal Adversarial Perturbations with Generative Models. 1st Deep Learning and Security Workshop 2018.
  • Slides

  • Invited talk on Adversarial Machine Learning. UCL Information Security Seminar, IBM Research, Microsoft Research, Facebook (2018).
  • TASP: Towards Anonymity Sets that Persist. WPES 2016, Satellite Workshop of CCS 2016, October, 2016, Vienna, Austria
  • k-fingerprinting: a Robust Scalable Website Fingerprinting Technique. USENIX Security 2016, August, 2016, Austin, USA.
  • Slides

  • Traffic Confirmation Attacks Despite Noise. Understanding and Enhancing Online Privacy Satellite Workshop of NDSS, February 21, 2016, San Diego, USA.
  • Slides

  • Guard Sets for Onion Routing. Proceedings on Privacy Enhancing Technologies 2015. Philadelphia, USA.
  • Guard Sets for Onion Routing. UCL Infosec Seminar 2015. London, UK.
  • Slides

  • Parameterized complexity of Alliances in Graphs. Programming, Logic, and Semantics Reading Group. Cambridge, UK.
  • Secure Sets in Graphs. Programming, Logic, and Semantics Reading Group, Computer Lab, University of Cambridge, Cambridge, December, 2013.

Misc. & Side Projects


    For a new project I need to keep fine-grained information about a browsing session. I couldn't find a suitable tool so I created one myself. Updates will continue to be pushed as it is in the early stages of development. Follow them here.
    August 05, 2016


    I am making available some of the less sensitive Numerai code here.
    July 07, 2016

    Generative approaches to unsupervised representation learning

    I trained a GAN to learn what a face looks like and then produce its own examples. Read more about that here.
    July 01, 2016

    Decrypt WLAN traffic

    I wrote a small educational guide of the dangers of browsing without HTTPS. Read more about that here.
    July 01, 2016

    An(other) interesting property of the Mandelbrot set

    June 01, 2015