Ambar Pal

About Me

I am a PhD student in the Computer Science Department at the Johns Hopkins University. I am advised by René Vidal and also work with Jeremias Sulam. I am affiliated with the Mathematical Institute for Data Science and the Vision Lab at JHU. Previously I obtained my Bachelor's degree in Computer Science from IIIT Delhi. My CV can be found here.

In the past, I have been fortunate to have worked with Chetan Arora[IIT Delhi], Gaurav Sharma[NEC Labs], H.B. Acharya[RIT], Somitra Sanadhya[IIT Ropar], Rajiv Raman[IIIT Delhi] and Yan Liu[USC].

Research Interests

The central theme of my research is understanding why deep learning works so well, and what are some theoretical explanations behind its successes and failures. With this aim, I have firstly worked on understanding common training time tricks like Dropout and Batch-Normalization. Secondly, I have worked on understanding and improving adversarial attacks and defenses for neural networks, by formulating a game-theoretic framework to study them, as well as exploring principled ways to defend against adversarial examples for the data domains of Graphs and Videos.

I received the JHU MINDS Data Science Fellowship in 2020 and 2019. My research is also supported by NSF and DARPA grants.

Publications

• A Game Theoretic Analysis of Additive Adversarial Attacks and Defenses [NeurIPS 2020]
  Ambar Pal, René Vidal.


• On the Regularization Properties of Structured Dropout [CVPR 2020]
  Ambar Pal, Connor Lane, René Vidal, Benjamin D. Haeffele.


• Making Deep Neural Network Fooling Practical [ICIP 2018]
  Ambar Pal, Chetan Arora.


• An Empirical Evaluation of Visual Question Answering for Novel Objects [CVPR 2017]
  Santhosh Kumar R, Ambar Pal, Gaurav Sharma, Anurag Mittal.

News

• [Oct 20] Our paper on developing a game-theoretic framework for additive adversarial attacks and defenses is accepted at NeurIPS 2020! Check out the paper here
• [Sep 20] Our work on adversarial learning is accepted for presentation at DeepMath 2020
• [June 20] I will be speaking at the Pontifical Catholic University of Chile on Structured Dropout.
• [June 20] I will be speaking at the JHU MINDS seminar on Structured Dropout regularization.
• [Feb 20] Our work on understanding structured dropout regularization is accepted at CVPR 2020! Find the paper here
• [Dec 19] I am a recipient of the JHU MINDS Fellowship!
• [Dec 19] I will be speaking at IIT Delhi at the Computer Science Department .
• [Nov 19] I will be at the NYU DeepMath Conference . Check out our poster on Structured Dropout.
• [May 19] I will be at UCLA at the workshop on Geometry in Deep Learning, and later at the Simons Insitute at UC Berkeley for the Deep Learning BootCamp.
• [April 19] I will be at the JHU Workshop on Computer Vision Applications.
• [March 19] I will be at ICERM, Brown University for the Theory and Practice in Machine Learning and Computer Vision Workshop.
• [September 18] I joined JHU as a PhD Student in the CS Department.
• [May 18] Our work on Practical Fooling of Deep Neural Networks is accepted at ICIP 18! Find the paper here.
• [July 17] I will be at CVPR 2017, being held at Honolulu, Hawaii in the last week of July.
• [March 17] Our work on Visual Question Answering for Novel Objects is accepted at CVPR 2017! Find the paper here.
• [Jan 17] I am selected for the IUSSTF-Viterbi India Program. I will be working under Yan Liu at the USC Melady Lab.
• [May 16] I will be at IIT Kanpur as a Student Research Associate at the Computer Vision Lab for summer. I am advised by Gaurav Sharma.

Preprints

• Cells in the Internet of Things
  Ayush Shah, H. B. Acharya, Ambar Pal.


• The Internet of Things: Perspectives on Security from RFID and WSN
  Ayush Shah, Ambar Pal, H. B. Acharya.