Shivam Duggal

I am a Ph.D. student at the Massachusetts Institute of Technology (MIT) EECS Department, working with Prof. Phillip Isola, Prof. Antonio Torralba and Prof. Bill Freeman.

Previously, I was a master's student at Carnegie Mellon University's Robotics Institute (MSR), advised by Prof. Deepak Pathak. I was fortunate to be a Siebel Scholar at CMU. Before CMU, I worked as a Research Scientist at Uber ATG Toronto, advised by Prof. Raquel Urtasun. I also worked closely with Prof. Shenlong Wang.

My research interests lie at the intersection of computer vision, machine learning, and robotics, with a focus on designing intelligent algorithms. More details later.

Email | Google Scholar | Github | Twitter

Recent Highlights

	•	Awarded Amazon AI PhD Fellowship 2025-2026. [MIT News]
	•	Started Ph.D. at Massachusetts Institute of Technology (MIT) EECS department, Fall 2023.
	•	Research Intern at Google Cambridge, Summer 2023.
	•	Graduated from CMU in May 2023. Special thanks to my amazing advisor Deepak Pathak. [Thesis]
	•	Awarded Siebel Scholarship 2023 ($35000 USD). [CMU News / Siebel Foundation News]
	•	Recognized as an Outstanding Reviewer at ECCV 2022.
	•	Our paper GeoSim was nominated for the Best Paper Award at CVPR 2021.
	•	ACM ICPC India Finalist: National Rank 28.

Publications

Single-pass Adaptive Image Tokenization for Minimum Program Search

Shivam Duggal, S. Byun, William T. Freeman, Antonio Torralba, Phillip Isola

NeurIPS 2025 (San Diego, USA)

paper | code | tl;dr (tweet)

Keywords: Algorithmic Information Theory, Kolmogorov Complexity, Upside-Down RL, Compression, Representation Learning.

Adaptive Length Image Tokenization via Recurrent Allocation

Shivam Duggal, Phillip Isola, Antonio Torralba, William T. Freeman

ICLR 2025 (Singapore)

paper | code | tl;dr (tweet)

Keywords: Recurrent Tokenizer, Compression, Representation Learning.

Your Diffusion Model is Secretly a Zero-Shot Classifier

Alexander Li, Mihir Prabhudesai, Shivam Duggal, Ellis Brown, Deepak Pathak

ICCV 2023 (Paris, France)

paper | project page | code

Topologically-Aware Deformation Fields for Single-View 3D Reconstruction

Shivam Duggal, Deepak Pathak

CVPR 2022 (New Orleans, USA)

paper | project page | code | video

Mending Neural Implicit Modeling for 3D Vehicle Reconstruction in the Wild

Shivam Duggal*, Zihao Wang*, Wei-Chiu Ma, Sivabalan Manivasagam, Justin Liang, Shenlong Wang, Raquel Urtasun

WACV 2022 (Hawaii, USA)

paper | video | poster

GeoSim: Realistic Video Simulation via Geometry-Aware Composition for Self-Driving (Oral)

Yun Chen*, Frieda Rong*, Shivam Duggal*, Shenlong Wang, Xinchen Yan, Sivabalan Manivasagam, Shangjie Xue, Ersin Yumer, Raquel Urtasun.

CVPR 2021 (Virtual)
(Oral, Best Paper Finalist)

paper | project page | video | poster

DeepPruner: Learning Efficient Stereo Matching via Differentiable PatchMatch

Shivam Duggal, Shenlong Wang, Wei-Chiu Ma, Rui Hu, Raquel Urtasun

ICCV 2019 (Seoul, Korea)
(Mentioned in Rick Szeliski's book, Depth Estimation Chapter)

paper | project page | code

Online Video Summarization: Predicting Future to Better Summarize Present

Shamit Lal*, Shivam Duggal*, Indu Sreedevi

WACV 2019 (Hawaii, USA)

paper

Other Projects

NeuralODF: Learning Omnidirectional Distance Fields for 3D Shape Representation

Trevor Houchens, Cheng-You Lu, Shivam Duggal, Rao Fu, Srinath Sridhar

arXiv 2022

Non-Exponential Radiative Transfer for Volumetric Light Transport

Shivam Duggal

CMU 15-668 Physics-based Rendering (Spring 2022)

report

Miscellaneous

I am a sports enthusiast and particularly enjoy playing cricket, badminton, volleyball, soccer and biking. I currently play for the MIT Cricket team as well as a club in Boston. Here are some old memories: here, here, here, here !!