Papers
24
Total Citations
1,646
H-Index
13
About
Abhinav Gupta is a pioneering robotics and machine learning researcher whose work sits at the intersection of robot learning, visual imitation, and large-scale data-driven manipulation. He is perhaps best known for his landmark 2016 paper "Supersizing Self-Supervision," which challenged the reliance on human-labeled datasets by demonstrating that robots could autonomously learn grasping through 50,000 real-world attempts and 700 hours of self-directed practice — a paper that has since accumulated over 1,000 citations and helped define the paradigm of self-supervised robot learning. Gupta has consistently pushed toward making robot learning practical beyond controlled lab settings, as reflected in his work on home environments, visual imitation from unstructured human videos (WHIRL), and large-scale demonstration datasets like MIME and DROID. His development of PyRobot, an open-source robotics framework, further demonstrates his commitment to democratizing robotics research infrastructure. More recently, his investigations into dexterous manipulation and motor program discovery signal a deepening focus on generalizable, compositional robot behavior. Across more than a decade of contributions, Gupta's research has fundamentally shaped how robots learn from data at scale.
Research Focus
Key Achievements
Top Papers
- 1Supersizing self-supervision: Learning to grasp from 50K tries and 700 robot hours1,099 citations · 2016
- 2DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset108 citations · 2024
- 3Human-to-Robot Imitation in the Wild77 citations · 2022
- 4Robot Learning in Homes: Improving Generalization and Reducing Dataset Bias61 citations · 2018
- 5PyRobot: An Open-source Robotics Framework for Research and Benchmarking51 citations · 2019
- 6
- 7
- 8Discovering Motor Programs by Recomposing Demonstrations22 citations · 2020
- 9
- 10Visual Imitation Made Easy22 citations · 2020