Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Liu, Zhirui, Ji, Kaiyang, Yang, Ke, Fan, Yahao, Yu, Jingyi, Shi, Ye, Wang, Jingya
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2511.22963
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911668320075776
author Liu, Zhirui
Ji, Kaiyang
Yang, Ke
Fan, Yahao
Yu, Jingyi
Shi, Ye
Wang, Jingya
author_facet Liu, Zhirui
Ji, Kaiyang
Yang, Ke
Fan, Yahao
Yu, Jingyi
Shi, Ye
Wang, Jingya
contents Enabling humanoid robots to follow free-form natural language commands is a critical step toward seamless human-robot interaction and general-purpose embodied AI. However, existing methods remain limited, often constrained to simple instructions or forced to sacrifice motion diversity for physical plausibility. To address this gap, we present Humanoid-LLA, a Large Language Action model that translates unconstrained natural language directly into executable whole-body motions for humanoid robots. Our approach tackles two core challenges: paired language-humanoid motion data scarcity and physical instability. First, we bridge high-level language semantics with physically-grounded control by learning a unified human-humanoid motion vocabulary. Second, we introduce a novel two-stage fine-tuning framework that begins with supervised motion Chain-of-Thought learning, followed by reinforcement learning refined with physical feedback to ensure robustness and stability. Extensive evaluation in simulation and real-world cross-embodiment experiments demonstrates that Humanoid-LLA achieves superior generalization to novel language commands and diverse motion generation while maintaining high physical fidelity.
format Preprint
id arxiv_https___arxiv_org_abs_2511_22963
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Commanding Humanoid by Free-form Language: A Large Language Action Model with Unified Motion Vocabulary
Liu, Zhirui
Ji, Kaiyang
Yang, Ke
Fan, Yahao
Yu, Jingyi
Shi, Ye
Wang, Jingya
Robotics
Artificial Intelligence
Enabling humanoid robots to follow free-form natural language commands is a critical step toward seamless human-robot interaction and general-purpose embodied AI. However, existing methods remain limited, often constrained to simple instructions or forced to sacrifice motion diversity for physical plausibility. To address this gap, we present Humanoid-LLA, a Large Language Action model that translates unconstrained natural language directly into executable whole-body motions for humanoid robots. Our approach tackles two core challenges: paired language-humanoid motion data scarcity and physical instability. First, we bridge high-level language semantics with physically-grounded control by learning a unified human-humanoid motion vocabulary. Second, we introduce a novel two-stage fine-tuning framework that begins with supervised motion Chain-of-Thought learning, followed by reinforcement learning refined with physical feedback to ensure robustness and stability. Extensive evaluation in simulation and real-world cross-embodiment experiments demonstrates that Humanoid-LLA achieves superior generalization to novel language commands and diverse motion generation while maintaining high physical fidelity.
title Commanding Humanoid by Free-form Language: A Large Language Action Model with Unified Motion Vocabulary
topic Robotics
Artificial Intelligence
url https://arxiv.org/abs/2511.22963