
David Held
Associate Professor · Robotics Institute — Faculty · Carnegie Mellon University
At Prof Rod’s reading desk
Questions to explore
- How should a robot choose where to make contact with an unfamiliar object?
- What evidence distinguishes simulation performance from transfer to a physical robot?
Research interests
David Held studies perceptual robot learning, connecting perception with the actions a robot can take. His work includes manipulating unfamiliar objects.
- Robot learning
- Computer vision
- Robot manipulation
Explore the work
Prof Rod’s reading notes
Start with a paper
CoRL · 2023
HACMan: Learning Hybrid Actor-Critic Maps for 6D Non-Prehensile Manipulation
Prof Rod’s reading note: Inspect the hybrid action representation before reading the success rates. Keep object generalization and simulation-to-real transfer as separate tests.
Bring a question of your own
Turn reading into research
Separate contact location from motion parameters. Design one test of unseen objects and another of transfer from simulation to a physical robot.
Keep exploring
More at Carnegie Mellon University
- Aaron Steinfeld
Robotics
- Abhinav Gupta
Machine learning · Computer vision
- Andrea Bajcsy
Human-robot interaction · Robot safety
- Cameron Riviere
Robotics
Improve this page
Suggest a correction, share a paper to start with, or point readers to a group page or opening.