Being-H0.5: Scaling Human-Centric Robot Learning for Cross-Embodiment Generalization

• URL: http://arxiv.org/abs/2601.12993v1
• Date: Mon, 19 Jan 2026 12:20:38 GMT
• Title: Being-H0.5: Scaling Human-Centric Robot Learning for Cross-Embodiment Generalization
• Abstract summary: We introduce Being-H0.5, a foundational Vision-Language-Action model designed for robust cross-embodiment generalization across diverse robotic platforms.<n>We present UniHand-2.0, the largest embodied pre-training recipe to date, comprising over 35,000 hours of multimodal data across 30 distinct robotic embodiments.<n>Being-H0.5 achieves state-of-the-art results on simulated benchmarks, such as LIBERO (98.9%) and RoboCasa (53.9%)
• Score: 38.20385682344082
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We introduce Being-H0.5, a foundational Vision-Language-Action (VLA) model designed for robust cross-embodiment generalization across diverse robotic platforms. While existing VLAs often struggle with morphological heterogeneity and data scarcity, we propose a human-centric learning paradigm that treats human interaction traces as a universal "mother tongue" for physical interaction. To support this, we present UniHand-2.0, the largest embodied pre-training recipe to date, comprising over 35,000 hours of multimodal data across 30 distinct robotic embodiments. Our approach introduces a Unified Action Space that maps heterogeneous robot controls into semantically aligned slots, enabling low-resource robots to bootstrap skills from human data and high-resource platforms. Built upon this human-centric foundation, we design a unified sequential modeling and multi-task pre-training paradigm to bridge human demonstrations and robotic execution. Architecturally, Being-H0.5 utilizes a Mixture-of-Transformers design featuring a novel Mixture-of-Flow (MoF) framework to decouple shared motor primitives from specialized embodiment-specific experts. Finally, to make cross-embodiment policies stable in the real world, we introduce Manifold-Preserving Gating for robustness under sensory shift and Universal Async Chunking to universalize chunked control across embodiments with different latency and control profiles. We empirically demonstrate that Being-H0.5 achieves state-of-the-art results on simulated benchmarks, such as LIBERO (98.9%) and RoboCasa (53.9%), while also exhibiting strong cross-embodiment capabilities on five robotic platforms.

Related papers

• X-WBC: A Cross-Embodiment Foundation Model for Humanoid Whole-Body Control [9.11598874129235]
  Scaling humanoid whole-body control requires large human motion corpora and training experience shared across robot bodies.<n>We introduce X-WBC, a cross-embodiment foundation framework that separates relatively shared human motion semantics from embodiment-specific physical execution.
  arXiv  Detail & Related papers  (2026-09-14T08:34:48Z)
• BRIDGE: An Open-Source Humanoid Platform via Morphology-Control Co-Design for Physical AI [38.7271333345027]
  Bridge is an open-source, 88cm-tall humanoid platform released alongside its control policy.<n>We demonstrate that Bridge captures human motion data with superior fidelity, exhibiting exceptional performance across locomotion, robust balance, and highly dynamic maneuvers.
  arXiv  Detail & Related papers  (2026-09-03T08:01:19Z)
• RLDX-1 Technical Report [74.70437517338186]
  We introduce RLDX-1, a general-purpose robotic policy for dexterous manipulation built on the Multi-Stream Action Transformer (MSAT)<n>Through empirical evaluation, we show that RLDX-1 consistently outperforms recent frontier Vision-Language-Action models.
  arXiv  Detail & Related papers  (2026-05-05T01:40:15Z)
• JoyAI-RA 0.1: A Foundation Model for Robotic Autonomy [90.77129709149574]
  JoyAI-RA is a vision-language-action embodied foundation model tailored for generalizable robotic manipulation.<n>JoyAI-RA bridges embodiment gaps, particularly between human manipulation and robotic control.<n>It outperforms state-of-the-art methods in both simulation and real-world benchmarks.
  arXiv  Detail & Related papers  (2026-04-22T01:51:48Z)
• FRoM-W1: Towards General Humanoid Whole-Body Control with Language Instructions [147.04372611893032]
  We present FRoM-W1, an open-source framework designed to achieve general humanoid whole-body motion control using natural language.<n>We extensively evaluate FRoM-W1 on Unitree H1 and G1 robots.<n>Results demonstrate superior performance on the HumanML3D-X benchmark for human whole-body motion generation.
  arXiv  Detail & Related papers  (2026-01-19T07:59:32Z)
• CEI: A Unified Interface for Cross-Embodiment Visuomotor Policy Learning in 3D Space [7.985647150826696]
  Cross-Embodiment Interface (CEI) is a framework for cross-embodiment learning.<n>CEI aligns robot trajectories through gradient-based optimization, followed by synthesizing observations and actions for unseen robot arms and end-effectors.<n>In experiments, CEI transfers data and policies from a Franka Panda robot to textbf16 different embodiments across textbf3 tasks in simulation.
  arXiv  Detail & Related papers  (2026-01-14T05:07:11Z)
• InternVLA-A1: Unifying Understanding, Generation and Action for Robotic Manipulation [77.07565723756119]
  We present InternVLA-A1, a vision-language-action model with dynamic predictive capabilities.<n>We pre-train these models on heterogeneous data sources over real-world robot data, synthetic simulation data, and human videos.<n>We evaluate InternVLA-A1 on 12 real-world robotic tasks and a simulation benchmark.
  arXiv  Detail & Related papers  (2026-01-05T18:54:29Z)
• UniAct: Unified Motion Generation and Action Streaming for Humanoid Robots [27.794309591475326]
  A long-standing objective in humanoid robotics is the realization of versatile agents capable of following diverse multimodal instructions with human-level flexibility.<n>Here we show that UniAct, a two-stage framework integrating a fine-tuned MLLM with a causal streaming pipeline, enables humanoid robots to execute multimodal instructions with sub-500 ms latency.<n>This approach yields a 19% improvement in the success rate of zero-shot tracking of imperfect reference motions.
  arXiv  Detail & Related papers  (2025-12-30T16:20:13Z)
• MiVLA: Towards Generalizable Vision-Language-Action Model with Human-Robot Mutual Imitation Pre-training [102.850162490626]
  We propose MiVLA, a vision-language-action model empowered by human-robot mutual imitation pre-training.<n>We show that MiVLA achieves strong improved generalization capability, outperforming state-of-the-art VLAs.
  arXiv  Detail & Related papers  (2025-12-17T12:59:41Z)
• UniBYD: A Unified Framework for Learning Robotic Manipulation Across Embodiments Beyond Imitation of Human Demonstrations [35.77665515297785]
  In embodied intelligence, the embodiment gap between robotic and human hands brings significant challenges for learning from human demonstrations.<n>We propose UniBYD, a unified framework that uses a dynamic reinforcement learning algorithm to discover manipulation policies aligned with the robot's physical characteristics.<n>To evaluate UniBYD, we propose UniManip, the first benchmark for cross-embodiment manipulation spanning diverse robotic morphologies.
  arXiv  Detail & Related papers  (2025-12-12T14:48:06Z)
• CEDex: Cross-Embodiment Dexterous Grasp Generation at Scale from Human-like Contact Representations [53.37721117405022]
  Cross-embodiment dexterous grasp synthesis refers to adaptively generating and optimizing grasps for various robotic hands.<n>We propose CEDex, a novel cross-embodiment dexterous grasp synthesis method at scale.<n>We construct the largest cross-embodiment grasp dataset to date, comprising 500K objects across four types with 20M total grasps.
  arXiv  Detail & Related papers  (2025-09-29T12:08:04Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.