Papers
15
Total Citations
1,410
H-Index
10
About
Justin Kerr is a leading researcher at the intersection of computer vision, graphics, and robotics, best known for his foundational work on Neural Radiance Fields (NeRF) and their application to real-world manipulation. He is the lead author of **Nerfstudio** (528 citations), a modular PyTorch framework that has become the de facto standard for NeRF development, dramatically accelerating research by providing a plug-and-play ecosystem. Kerr’s work bridges the gap between 3D scene representation and robotic perception. His **LERF** (291 citations) pioneered grounding natural language queries directly into 3D radiance fields, enabling robots to understand and interact with environments based on semantic descriptions. Demonstrating a knack for tackling notoriously hard problems, he developed **Dex-NeRF** (61 citations) to enable robotic grasping of transparent objects—a challenge that stymies conventional depth sensors. Earlier, his **PRIMAL** (399 citations) advanced multi-agent pathfinding using reinforcement and imitation learning, addressing scalability in warehouse automation. Kerr’s portfolio also includes innovative work in surgical robotics (learning to handle unmodified needles) and autonomous cable untangling, showcasing his ability to translate cutting-edge 3D representations into practical robotic solutions. His research consistently pushes the boundaries of how neural fields can empower robots to perceive and manipulate complex, unstructured environments.
Research Focus
Key Achievements
Top Papers
- 1Nerfstudio: A Modular Framework for Neural Radiance Field Development528 citations · 2023
- 2PRIMAL: Pathfinding via Reinforcement and Imitation Multi-Agent Learning399 citations · 2019
- 3LERF: Language Embedded Radiance Fields291 citations · 2023
- 4Dex-NeRF: Using a Neural Radiance Field to Grasp Transparent Objects61 citations · 2021
- 5Learning to Localize, Grasp, and Hand Over Unmodified Surgical Needles30 citations · 2022
- 6Autonomously Untangling Long Cables20 citations · 2022
- 7Nerfstudio: A Modular Framework for Neural Radiance Field Development19 citations · 2023
- 8
- 9
- 10SGTM 2.0: Autonomously Untangling Long Cables using Interactive Perception12 citations · 2023