Ankit Goyal

Nvidia (United States), Princeton University

Papers

10

Total Citations

732

H-Index

7

About

Ankit Goyal is a robotics and AI researcher whose work sits at the intersection of robot manipulation, task planning, and spatial reasoning. He is best known for developing **ProgPrompt**, a landmark framework that leverages large language models to generate programmatic, situationally aware task plans for robots — a contribution that has garnered over 600 citations across its versions and established him as a notable voice in LLM-driven robotics. His research on 3D object manipulation has produced influential systems including **RVT** and **RVT-2**, which use multi-view transformers to enable precise robotic manipulation from few demonstrations, and **IFOR**, an end-to-end approach to object rearrangement using iterative flow minimization. Goyal has also advanced spatial understanding through the **Rel3D** benchmark and explored 3D multiview pretraining for manipulation with **3D-MVP**. Earlier in his career, he applied computer vision to construction safety, developing methods for detecting hazardous human-robot interactions on job sites. Across his body of work, Goyal consistently bridges perception, language, and action — pushing robots closer to operating effectively in complex, unstructured real-world environments.

Research Focus

Key Achievements

7
H-Index
10
Papers
732
Total Citations
73
Avg Citations/Paper
🏆 Most Cited Paper
ProgPrompt: Generating Situated Robot Task Plans using Large Language Models
508 citations · 2023
📈 Most Prolific Year: 2023 (3 Papers)
🤝 Key Collaborators: 27
🏛 Institutions: Nvidia (United States), Princeton University

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago