PGDG generates diverse, successful recovery trajectories from a single demonstration using iterative physics-grounded sampling and zero-shot curation to improve bimanual policy robustness.
Real2render2real: Scaling robot data without dynamics simulation or robot hardware
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
PGDG: Physically Grounded Data Generation for Robust Bimanual Policy Learning from a Single Demonstration
PGDG generates diverse, successful recovery trajectories from a single demonstration using iterative physics-grounded sampling and zero-shot curation to improve bimanual policy robustness.