Varun Chandrasekaran

This page is no longer being maintained. I graduated from the Department of Computer Sciences at the University of Wisconsin-Madison, where I was fortunate to work with Suman Banerjee and Somesh Jha.
My research interests lie at the intersection of Security & Privacy and various domains such as Systems, Networking, Distributed Computing, and Machine Learning. In particular, my research aims to understand what private information can be inferred through interaction with a machine learning model.
Previously, I obtained my MS degree from the Courant Institute of Mathematical Sciences. I've obtained my B.Eng in Computer Science and Engineering from the College of Engineering, Guindy.

News

I will be joining the University of Illinois Urbana-Champaign as an Assistant Professor in Fall 2023, and am looking for motivated PhD students (in both ECE and CS).
Our group at Microsoft Research is also looking for interns.
If interested, please first read this and kindly follow the instructions. I will reach out if there's a fit.

Selected Publications

A General Framework For Detecting Anomalous Inputs to DNN Classifiers [arXiv][talk]

In the Proceedings of the 38th International Conference on Machine Learning (ICML), Long Presentation, 2021

Proof-of-Learning: Definitions and Practice [arXiv][talk]

In the Proceedings of the 42nd IEEE Symposium of Security & Privacy, 2021

Face-Off: Adversarial Face Obfuscation [arXiv][talk]

In the Proceedings of the 21st Privacy Enhancing Technologies Symposium (PETS), 2021

Machine Unlearning [arXiv][talk]

In the Proceedings of the 42nd IEEE Symposium of Security & Privacy, 2021

Exploring Connections Between Active Learning and Model Extraction [arXiv][talk]

In the Proceedings of the 29th USENIX Security Symposium, 2020

Conference Publications

[C11] Unrolling SGD: Understanding Factors Influencing Machine Unlearning [arXiv]
In the Proceedings of the 7th IEEE European Symposium of Security & Privacy, 2022
[C10] CONFIDANT: A Privacy Controller for Social Robots [arXiv]
In the Proceedings of the 17th ACM/IEEE International Conference on Human-Robot Interaction (HRI), 2022
[C9] PowerCut and Obfuscator: An Exploration of the Design Space for Privacy-Preserving Interventions for Voice Assistants [arXiv][talk]
In the Proceedings of the 17th USENIX Symposium on Usable Privacy and Security (SOUPS), 2021
[C8] A General Framework For Detecting Anomalous Inputs to DNN Classifiers [arXiv][talk]
In the Proceedings of the 38th International Conference on Machine Learning (ICML), Long Presentation, 2021
[C7] Proof-of-Learning: Definitions and Practice [arXiv][talk]
In the Proceedings of the 42nd IEEE Symposium of Security & Privacy, 2021
[C6] Entangled Watermarks as a Defense against Model Extraction [arXiv][talk]
In the Proceedings of the 30th USENIX Security Symposium, 2021
[C5] Face-Off: Adversarial Face Obfuscation [arXiv][talk]
In the Proceedings of the 21st Privacy Enhancing Technologies Symposium (PETS), 2021
[C4] Machine Unlearning [arXiv][talk]
In the Proceedings of the 42nd IEEE Symposium of Security & Privacy, 2021
Press: IEEE Spectrum
[C3] Exploring Connections Between Active Learning and Model Extraction [arXiv][talk]
In the Proceedings of the 29th USENIX Security Symposium, 2020
[C2] A Framework for Analyzing Spectrum Characteristics in Large Spatio-temporal Scales [pdf]
In the Proceedings of the 24th ACM MobiCom, 2019
[C1] Alphacodes: Usable, Secure Transactions with Untrusted Providers using Human Computable Puzzles [pdf]
In the Proceedings of the 7th ACM DEV, 2016

Workshop Publications

  1. [W3] Analyzing And Improving Neural Networks By Generating Semantic Counterexamples Through Differentiable Rendering [arXiv]
    In the Uncertainty & Robustness in Deep Learning (UDL) Workshop at ICML, 2021
  2. [W2] Causally Constrained Data Synthesis For Private Data Release [pdf]
    In the Distributed and Private Machine Learning (DPML) Workshop at ICLR, 2021
  3. [W1] Traversing the Quagmire that is Privacy in your Smart-Home [pdf]
    In the Proceedings of ACM SIGCOMM Workshop on IoT Security and Privacy, 2018

Manuscripts

[P5] SoK: Machine Learning Governance [arXiv]
[P4] On the Exploitability of Audio Machine Learning Pipelines to Surreptitious Adversarial Examples[arXiv]
[P3] Causally Constrained Data Synthesis For Private Data Release [arXiv]
[P2] On the Effectiveness of Mitigating Data Poisoning Attacks with Gradient Shaping [arXiv]
[P1] Rearchitecting Classification Frameworks For Increased Robustness [arXiv]

Instruction Experience

  • Networks and Mobile Systems (CSCI-GA.2620-001), Spring 2016

  • Computer System Organisation (CSCI-UA.0201-002), Fall 2015

  • Technology and Economic Development (CSCI-UA.0380-003), Spring 2015
  • Feel free to contact me

    The best way to contact me is through e-mail.

    Email

    chandrasekaran [at] cs.wisc.edu

    Address

    1210 West Dayton Street
    Madison, WI 53706