Close

Vivek Sivaraman Narayanaswamy


Machine Learning Research Scientist

Lawrence Livermore National Laboratory

Download CV

About Me

Vivek photo

Hello and welcome to my website! I'm Vivek, a Machine Learning Research Scientist at LLNL, Livermore, California. I have a strong publication record (NeurIPS, ICML, ECCV, ICCV, AAAI, Interspeech, ICASSP, IEEE Access, IDT) and hands-on experience solving problems in computer vision, speech/audio, scientific machine learning, interpretability, and healthcare AI.

Previously, I was a postdoctoral researcher at LLNL, mentored by Dr. Jayaraman J. Thiagarajan, where I worked on a wide range of topics including surrogate modeling, failure detection in deep models, out-of-distribution generalization, robustness, and uncertainty quantification.

I completed my Ph.D. in Electrical Engineering at Arizona State University under the supervision of Dr. Andreas Spanias and Dr. Jayaraman J. Thiagarajan. My doctoral research focused on building reliable and deployable deep learning algorithms. Curious? You can read my thesis here.

My core research interests include (but are not limited to): deep learning, supervised and unsupervised learning, generative modeling, inverse problems, uncertainty quantification, explainable AI, and out-of-distribution detection.

If you're interested in collaborating, feel free to reach out via email — I’m usually very responsive!

News

  • April 2025 - Paper titled ;The Surprising Utility of Group Partitioning in Improving Conformal Prediction of Visual Classifiers under Distributional Shifts' accepted to CVPR UnCV Workshop 2025 - Preprint coming soon!
  • April 2025 - Paper titled 'Leveraging Registers in Vision Transformers for Robust Adaptation' accepted as an oral at IEEE ICASSP 2025 - Link
  • January 2025 - Accepted the ML Research Scientist Position at LLNL
  • October 2024 - Paper titled 'On the Use of Anchoring for Training Vision Models' accepted to NeurIPS 2024 as a spotlight poster (top 25%) - Link to Project Website

Experience

Lawrence Livermore National Laboratory

ML Research Scientist

Lawrence Livermore National Laboratory

Post-doctoral Research Scientist

Lawrence Livermore National Laboratory

Computing Scholar Intern

Lawrence Livermore National Laboratory

Computing Scholar Intern

Lawrence Livermore National Laboratory

Computing Scholar Intern

Qualcomm R&D

Interim Engineering Intern

Arizona State University

Graduate Research Associate

Education

Arizona State University

August 2017 - Jan 2023

Ph.D. in Electrical Engineering

Thesis: Methodologies to Improve Fidelity and Reliability of Deep Learning Models for Real-World Deployment

Courses: Digital Signal Processing, Communication Systems, Random Signal Theory, Detection and Estimation Theory, Adaptive Signal Processing, Speech Processing and Audio Perception, Statistical Machine Learning, Convex Optimization, Artificial Neural Computation.

Anna University, India

June 2013 - May 2017

Bachelor of Engineering in Electronics and Communication Engineering

Publications

Improved Medical Out-of-Distribution Detectors For Modality and Semantic Shifts

V. Narayanaswamy, Y.Mubarka, R. Anirudh, D. Rajan, A. Spanias, J. J. Thiagarajan, ICML2022 Principles of Distribution Shifts Workshop

Link

Using Direct Error Predictors to Improve Model Safety and Interpretability

V. Narayanaswamy, D. Rajan, A. Spanias, J. J. Thiagarajan, ICML2022 Principles of Distribution Shifts Workshop

Link

Improved StyleGAN-v2 based Inversion for Out-of-Distribution Images

V. Narayanaswamy, R. Subramayam, M. Naufel, A. Spanias, J. J. Thiagarajan, ICML 2022

Link

Designing Counterfactual Generators using Deep Model Inversion

J. J. Thiagarajan, V. Narayanaswamy, D. Rajan, J. Liang, A. Chaudhary, A. Spanias, Neurips 2021

Link

Accurate and Robust Feature Importance Estimation Under Distribution Shifts

J. J. Thiagarajan, V. Narayanaswamy, R. Anirudh, P. Bremer, AAAI 2021

Link

On the Design of Deep Priors for Unsupervised Audio Restoration

V. Narayanaswamy, J. J. Thiagarajan, A. Spanias, Interspeech 2021

Link

Using Deep Image Priors to Generate Counterfactual Explanations

V. Narayanaswamy, J. J. Thiagarajan, A. Spanias, ICASSP 2021

Link

Unsupervised Audio Source Separation using Generative Priors

V. Narayanaswamy, J. J. Thiagarajan, R. Anirudh, A. Spanias, Interspeech 2020

Link

Selected publications are listed above. For the complete list of publications, visit my Google Scholar!

Skills

Get in Touch


Email ID: narayanaswam1@llnl.gov