Research Scientist at Meta AI
satyanshukla@fb.com
Google Scholar
I am a Senior Research Scientist at Meta AI, where I specialize in scaling and improving multimodal models.
My work focuses on multimodal understanding, foundational models, video modeling and large scale training. I'm also interested in
making deep learning models robust to missing or incomplete data, uncertainty modeling, and adversarial examples.
Before joining Meta, I completed my Ph.D. in Computer Science at UMass Amherst advised by Benjamin Marlin. My
thesis focused on building
deep learning models for irregularly sampled and incomplete time series.
During my Ph.D., I spent wonderful summers interning at Microsoft Research, Bosch Center for AI, and Facebook.
I earned my Bachelors (Hons.) and Masters in Electrical Engineering from
Indian Institute of Technology Kharagpur and
was awarded the prestigious Institute Silver Medal.
Some of my work is available as preprints on arXiv. [* denotes equal contribution]
uCAP: An Unsupervised Prompting Method for Vision-Language Models
Tuan Nguyen, Kai Sheng Tai, Sirius Chen, Satya Narayan Shukla, Hanchao Yu, Philip Torr, Taipeng Tian, Ser-Nam Lim
European Conference on Computer Vision (Oral), 2024
Learning to Localize Objects Improves Spatial Reasoning in Visual-LLMs
Kanchana Ranasinghe, Satya Narayan Shukla, Omid Poursaeed, Michael Ryoo, Tsung-Yu Lin
Conference on Computer Vision and Pattern Recognition, 2024
The Belebele Benchmark: a Parallel Reading Comprehension Dataset in 122 Language Variants
Lucas Bandarkar, Davis Liang, Benjamin Muller, Mikel Artetxe, Satya Narayan Shukla, Donald Husa, Naman Goyal, Abhinandan Krishnan, Luke Zettlemoyer, Madian Khabsa
Association for Computational Linguistics (ACL), 2024
Revisiting Kernel Temporal Segmentation as an Adaptive Tokenizer for Long-form Video Understanding
Afham Aflal, Satya Narayan Shukla, Omid Poursaeed, Pengchuan Zhang, Ashish Shah, Sernam Lim
International Conference on Computer Vision Workshop, 2023
Universal Pyramid Adversarial Training for Improved ViT Performance
Ping Chiang, Yipin Zhou, Omid Poursaeed, Satya Narayan Shukla, Tom Goldstein, Sernam Lim
CoRR, abs/2312.16339, 2023
Heteroscedastic Temporal Variational Autoencoder For Irregularly Sampled Time Series
Satya Narayan Shukla, Benjamin Marlin
International Conference on Learning Representations, 2022
Deep Learning Models for Irregularly Sampled and Incomplete Time Series
Satya Narayan Shukla
Doctoral Dissertations, University of Massachusetts Amherst, 2021
Simple and Efficient Hard Label Black-box Adversarial Attacks in Low Query Budget Regimes
Satya Narayan Shukla, Anit Kumar Sahu, Devin Willmott, Zico Kolter
ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2021
Multi-Time Attention Networks for Irregularly Sampled Time Series
Satya Narayan Shukla, Benjamin Marlin
International Conference on Learning Representations, 2021
A Survey on Principles, Models and Methods for Learning from Irregularly Sampled Time Series
Satya Narayan Shukla, Benjamin Marlin
ML Retrospectives, Surveys & Meta-Analyses (ML-RSA) Workshop at NeurIPS, 2020
Gaussian MRF Covariance Modeling for Efficient Black-Box Adversarial Attacks
Anit Kumar Sahu, Satya Narayan Shukla, Zico Kolter
CoRR, abs/2010.04205, 2020
Integrating Physiological Time Series and Clinical Notes with Deep Learning for Improved ICU Mortality Prediction
Satya Narayan Shukla, Benjamin Marlin
ACM Conference on Health, Inference, and Learning, Workshop Track, 2020
Assessing the Adversarial Robustness of Monte Carlo and Distillation Methods for Deep Bayesian Neural Network Classification
Meet Vadera*, Satya Narayan Shukla*, Brian Jalaian, Benjamin Marlin
Artificial Intelligence Safety (SafeAI) Workshop at AAAI Conference on Artificial Intelligence, 2020
Black-box Adversarial Attacks with Bayesian Optimization
Satya Narayan Shukla, Anit Kumar Sahu, Devin Willmott, Zico Kolter
CoRR, abs/1909.13857, 2019
Interpolation-Prediction Networks for Irregularly Sampled Time Series
Satya Narayan Shukla, Benjamin Marlin
International Conference on Learning Representations, 2019
Modeling Irregularly Sampled Clinical Time Series
Satya Narayan Shukla, Benjamin Marlin
Machine Learning for Health (ML4H) Workshop at Neural Information Processing Systems, 2018
Prediction and Imputation in Irregularly Sampled Clinical Time Series Data using Hierarchical Linear Dynamical Models
Abhishek Sengupta, Prathosh AP, Satya Narayan Shukla, Vaibhav Rajan, Chandan K Reddy
39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 2017
Estimation of Blood Pressure from Non-invasive Data
Satya Narayan Shukla
39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 2017
Non-invasive Cuffless Blood Pressure Measurement by Vascular Transit Time
Satya Narayan Shukla, Karan Kakwani, Amit Patra, Bipin Kumar Lahkar, Vivek Kumar Gupta, Alwar Jayakrishna, Puneet Vashisht, Induja Sreekanth
IEEE International Conference on VLSI Design, 2015
Bayesian-optimization-based Query-efficient Black-box Adversarial Attacks
Satya Narayan Shukla, Anit Kumar Sahu, Devin Willmott, Zico Kolter
Methods and Systems for Modeling Irregularly Sampled Temporal Data using Kalman Filters
Abhishek Sengupta, Prathosh AP, Satya Narayan Shukla, Vaibhav Rajan, Katerina Sinclair, Stephen Fullerton
Forecasting Patient Vital Measurements for Healthcare Analytics
Abhishek Sengupta, Bhupendra Singh Solanki, Prathosh AP, Vaibhav Rajan, Katerina Sinclair, Stephen Fullerton, Satya Narayan Shukla