About me

I am a fourth year PhD student in the Department of Computer Science at University of Maryland, College Park. I am advised by Prof. Rama Chellappa and Prof. Soheil Feizi . My research interests are in the areas of Machine Learning and Computer Vision. More specifically, I work on robustness of neural networks to domain shifts and adversarial perturbation.

Previously, I obtained a Bachelor's degree (B.Tech), with a major in Electrical Engineering and a minor in Operations Research from Indian Institute of Technology, Madras.



  • Adversarial Robustness of Flow-Based Generative Models
    Phillip Pope * , Yogesh Balaji * , and Soheil Feizi, To appear in AISTATS 2020 New!

  • Instance adaptive adversarial training: Improved accuracy tradeoffs in neural nets
    Yogesh Balaji , Tom Goldstein and Judy Hoffman, Under review New!

  • Normalized Wasserstein Distance for Mixture Distributions with Applications in Adversarial Learning and Domain Adaptation
    Yogesh Balaji , Rama Chellappa and Soheil Feizi, ICCV 2019

  • Conditional GAN with Discriminative Filter Generation for Text-To-Video Synthesis
    Yogesh Balaji , Martin Renqiang Min, Bing Bai, Rama Chellappa and Hans Peter Graf, IJCAI 2019

  • Entropic GANs meet VAEs: A Statistical Approach to Compute Sample Likelihoods in GANs
    Yogesh Balaji , Hamed Hassani, Rama Chellappa and Soheil Feizi, To appear in ICML 2019

  • MetaReg: Towards Domain Generalization using Meta-Regularization
    Yogesh Balaji , Swami Sankaranarayanan and Rama Chellappa, NeurIPS 2018

  • Generate To Adapt: Unsupervised Domain Adaptation using Generative Adversarial Networks
    Swami Sankaranarayanan *, Yogesh Balaji *, Carlos D. Castillo and Rama Chellappa, CVPR 2018 ( Spotlight)

  • Learning from Synthetic Data: Addressing Domain Shift for Semantic Segmentation
    Swami Sankaranarayanan *, Yogesh Balaji *, Arpit Jain, Ser Nam Lim and Rama Chellappa, CVPR 2018 ( Spotlight )

  • Unrolling the Shutter: CNN to Correct Motion Distortions
    Vijay Rengarajan, Yogesh Balaji , and A.N. Rajagopalan, CVPR 2017 ( Oral paper )

  • Deep Decoupling of Defocus and Motion Blur for Dynamic Segmentation
    Abhijith Punnappurath, Yogesh Balaji , Mahesh Mohan, and A. N. Rajagopalan, ECCV 2016

    * Equal contribution

    Technical Reports

  • Deep Learning for estimating distortions in images
    B.Tech thesis, Indian Institute of Technology Madras, 2016


  • Facebook AI Research, Menlo Park, CA

    Topic: Adversarial robustness
    I spent the summer of 2019 interning at Facebook AI Research, where I worked on developing algorithms for improving robustness-accuracy trade-off in adversarial training.

  • NEC Laboratories America, Princeton, NJ

    Topic: Generative models for sequential data
    I spent the summer of 2018 at NEC Labs where I worked on designing text conditioned video generation models using Generative Adversarial Networks

  • Qualcomm Research and Development, San Diego

    Topic: Randomized Algorithms for Wireless Systems
    I spent the summer of 2015 at Qualcomm Research, San Diego. Here, I worked with Wireless Systems group on developing MAC layer protocols that would enable fair coexistence between WiFi and LTE-U (LTE in unlicensed spectrum)


email :   yogesh'at'cs.umd.edu

Address :   Room 4436, A.V. Williams Building,
                      University of Maryland,
                      College Park, MD 20742