I am a final-year Computer Science PhD Candidate at Harvard University and Boston Children's Hospital. I am advised by Hanspeter Pfister and Gabriel Kreiman. My research focusses on understanding and improving out-of-distribution generalization in Machine Learning and its applications to the Sciences.
Looking for Post Doc positions starting Jan 2025 at the intersection of Machine Learning and the Sciences
Research
Generalizing beyond the training data (i.e., out-of-distribution) is a fundamental goal in machine learning. While standardized datasets have led to unprecedented success, understanding how our models generalize beyond them is essential for building robust, trustable models that we can deploy with confidence. My research approaches this problem using computer graphics to create custom 3D worlds with complete control over the training and testing distributions, which we then use to investigate and improve out-of-distribution generalization. For an updated list of my papers, please check google scholar.
References
- Hanspeter Pfister: pfister@seas.harvard.edu
- Gabriel Kreiman: gabriel.kreiman@childrens.harvard.edu
- Fredo Durand: fredo@mit.edu
Students Advised
- Ravi Srinivasan, Master's. Now: PhD Student, UC Berkeley
- Serena Bono, Master's. Now: PhD Student, MIT
- Yash Gupta, Undergraduate. Now: Master's Student, MIT
- Arshika Lalan, Undergraduate. Now: Master's Student, CMU.
Beyond Research
- I am an avid guitarist. You can find my covers and original music here - Bandcamp and Soundcloud.
- Learning to Speak Python: A course I made during the pandemic to teach people from diverse backgrounds (lawyers, financial consultants, medical practioners) how to code in Python.
Reviewing Experience
- Journals: TPAMI, Pattern Recognition Letters, Neural Networks, TVCG
- Conferences: NeurIPS, ICML, IJCAI, AAAI, CVPR, ECCV, UIST, InfoViz.
