Shibani Santurkar
PhD Candidate
Cambridge, Massachusetts, United States
Summary
🤩
Rockstar🎓
Top SchoolShibani Santurkar is a PhD candidate at MIT CSAIL with a decade of experience in deep learning, specializing in unsupervised approaches for computer vision. Her research blends rigorous academic training (PhD at MIT, BTech/MTech from IIT Bombay) with hands-on engineering, including contributions to MadryLab's widely used robustness library where she added ImageNet and BREEDS dataset support and subpopulation-shift benchmarks. She has a background spanning energy-efficient circuit design and interactive graphics from internships at MIT and TU Braunschweig, giving her cross-domain perspective on systems and perceptual modeling. Based in Cambridge, MA, she focuses on making vision models more reliable and generalizable, pairing theoretical insight with practical dataset and tooling development.
10 years of coding experience
Doctor of Philosophy (PhD), Electrical Engineering and Computer Science, Doctor of Philosophy (PhD), Electrical Engineering and Computer Science at Massachusetts Institute of TechnologyIndian Institute of Technology Bombay
French, Hindi, English
Github Skills (11)
data-manipulation10
robust10
machine-learning10
python10
robustness10
imagenet10
datasets10
data-set10
pytorch9
data-science9
computer-vision8
Programming languages (2)
Github contributions (5)
MadryLab/robustness
Oct 2019 - Aug 2020
A library for experimenting with, training and evaluating neural networks, with a focus on adversarial robustness.
Role in this project:
ML Engineer
ML EngineerContributions:28 commits, 13 PRs, 31 pushes in 9 months
Contributions summary:Shibani primarily contributed to the development of the "robustness" library by implementing and integrating datasets related to ImageNet and BREEDS. Their work involved adding support for custom and open-image datasets, defining class hierarchies, and creating dataset split generators for subpopulation shift benchmarks. The commits demonstrate an understanding of data loading, manipulation, and integration within the context of adversarial robustness research. This user has also added supporting documentation.
pytorchexperimentingrobustnessdeep-learningadversarial
MadryLab/EditingClassifiers
Oct 2021 - Oct 2022
Contributions:9 commits, 10 pushes in 1 year
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.