multi-skill control 1perception 1quadrupedal locomotion 1reinforcement learning 1terrain navigation 1transformer models 1
From the 1 of 2 linked papers with an AI index.
2 papers
cs.RO2026
Agile perceptive multi-skill locomotion for quadrupedal robots in the wild
Jun-Gill Kang, Jaehyun Park, Tae-Gyu Song +3
The paper presents APT-RL, a transformer‑based reinforcement learning framework that learns multiple locomotion skills from simulated data and enables a quadrupedal robot to traver…
cs.RO2026
DynaFlow: Dynamics-embedded Flow Matching for Physically Consistent Motion Generation from State-only Demonstrations
Sowoo Lee, Dongyun Kang, Jaehyun Park +1
This paper introduces DynaFlow, a novel framework that embeds a differentiable simulator directly into a flow matching model. By generating trajectories in the action space and map…