Active Embodiment Identification with Reinforcement Learning for Legged Robots

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Bohlinger, Nico, Peters, Jan
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917473394098176
author Bohlinger, Nico
Peters, Jan
author_facet Bohlinger, Nico
Peters, Jan
contents We present an active embodiment identification method for legged robots that jointly learns information-seeking behavior and explicit embodiment prediction. Using a history-augmented URMA architecture, the method infers joint-level and global embodiment parameters through interaction with the environment in simulation across different morphologies.
format Preprint
id arxiv_https___arxiv_org_abs_2605_08020
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Active Embodiment Identification with Reinforcement Learning for Legged Robots
Bohlinger, Nico
Peters, Jan
Robotics
We present an active embodiment identification method for legged robots that jointly learns information-seeking behavior and explicit embodiment prediction. Using a history-augmented URMA architecture, the method infers joint-level and global embodiment parameters through interaction with the environment in simulation across different morphologies.
title Active Embodiment Identification with Reinforcement Learning for Legged Robots
topic Robotics
url https://arxiv.org/abs/2605.08020