2 papers
cs.RO2026
FCTS: Failure-Guided Curriculum Tree Search Over Generated Terrain for Legged Parkour
Narayanan Palghat Parameswaran, Dong-Ki Kim, Jay Patrikar +4
Quadruped parkour policies handle canonical obstacles well, but their training curricula often miss the irregular, compound geometry of industrial and disaster-response sites, wher…
cs.RO2026
MimicAgent: Quadruped Skills via Prompt-to-Trajectory Generation
Lucky Kant Nayak, Narayanan Palghat Parameswaran, Neehar Peri +1
We present MimicAgent, a prompt-to-trajectory generation framework for learning dynamic quadruped skills. Although reward shaping is extensively used when training quadruped polici…