Vikash Sehwag

vikash.png

About me

I am a PhD candidate in Electrical Engineering at Princeton University. I am co-advised by Prof. Prateek Mittal and Prof. Mung Chiang. Before joining Princeton, I completed my undergraduate at the Indian Institute of Technology, Kharagpur.

I am broadly interested in problems at the intersection of security and machine learning. I have significant expertise in the domain of adversarial attacks and defenses under a wide range of threat models. Under this direction, I have been working on improving both empirical and provable notions of the robustness of state-of-the-art deep neural networks.

News

Work in progress

Research projects

On Pruning Adversarially Robust Neural Networks
Vikash Sehwag, Shiqi Wang, Prateek Mittal, and Suman Jana
venue, 2020 (bibtex)
arxiv / poster / video / slides / demo / code / blog

We propsed a novel pruning approach which can compress neural networks upto 100x while achieving high adversarial robustness and accuracy.

Robust open-world machine learning
Vikash Sehwag, Arjun Nitin Bhagoji, Liwei Song, Chawin Sitawarin, Daniel Cullina, Mung Chiang, and Prateek Mittal
AISec, 2019 (bibtex)
arxiv / poster / video / slides / demo / code / blog

In this ongoing project we are rigourously analyzing the avdersarial robustness of open-world machine learning systems.