具身开发者社区

外观

论文SIMPLE: A Benchmark for Humanoid Loco-Manipulation18 天前

SIMPLE基准发布：融合MuJoCo与IsaacSim验证人形机器人零样本迁移 ​

针对人形机器人全身移动操作评测难题，SIMPLE基准整合MuJoCo动力学与IsaacSim渲染，提供16项任务及千级资产。实验证实仿真策略可零样本迁移至真机，为具身智能研究提供可扩展验证基础。

AI 整理
  • 仿真与训练
  • 数据与采集
  • 评测与安全

本文由 AI 根据公开内容整理，可能有误，请以原文为准：SIMPLE: Simulation-Based Policy Learning and Evaluation for Humanoid Loco-manipulation（SIMPLE: A Benchmark for Humanoid Loco-Manipulation）

核心内容 ​

针对人形机器人全身移动操作缺乏统一评测标准的痛点，研究团队发布SIMPLE仿真测试平台。该平台包含16项多样化全身任务、50个室内场景及超1000种物体资产，旨在解决现有基准局限于桌面或轮式机器人的问题，为人形基础模型提供可扩展且可复现的策略学习与评估环境。

技术上，SIMPLE耦合MuJoCo的高精度接触动力学与IsaacSim的光逼真渲染，集成运动规划自动轨迹生成与低延迟VR遥操作两条数据管线。基准测试覆盖轻量模仿网络、VLA大模型及世界动作模型（WAMs），实验表明仿真性能与真机强相关，且基于该平台数据训练的策略可实现零样本物理迁移。

关键看点 ​

  • 双引擎架构：耦合MuJoCo接触动力学与IsaacSim光逼真渲染，兼顾物理准确性与视觉真实性
  • 规模化资产：内置16项全身移动操作任务、50个室内场景及超1000种物体资产支撑泛化训练
  • 混合数据管线：集成运动规划自动生成轨迹与低延迟VR遥操作接口，加速高质量数据采集
  • 零样本迁移验证：证实仿真训练策略可直接部署至物理人形机器人，Sim-to-Real相关性显著

工程与产业观察 ​

SIMPLE通过统一动力学与渲染引擎并验证零样本迁移能力，确立了人形机器人全身操作Sim-to-Rel的新范式。其提供的标准化任务集与混合数据管线，有效降低了具身算法从仿真到真机的验证门槛，对VLA及世界模型在复杂接触场景下的落地具有重要参考价值。

英文摘要 (Abstract) ​

Humanoid foundation models are advancing faster than we can evaluate them. While real-world testing is expensive and difficult to reproduce, existing simulation benchmarks focus primarily on table-top or wheeled robots. A scalable and reproducible benchmark for whole-body humanoid loco-manipulation remains an open problem. To this end, we present SIMPLE, a unified simulation testbed for humanoid policy learning and evaluation. SIMPLE couples the accurate contact-rich dynamics of MuJoCo with the photorealistic rendering of IsaacSim. It provides a large-scale environment comprising 16 diverse whole-body tasks, 50 indoor scenes, and over 1,000 object assets. To facilitate scalable data collection, the framework integrates two data generation pipelines: automated trajectory generation via motion planning and a low-latency VR teleoperation interface. We further integrate and benchmark mainstream humanoid policies at scale in SIMPLE, including lightweight imitation networks, large vision-language-action (VLA) models, and recent world action models (WAMs). Our experiments reveal a strong correlation between policy performance in simulation and the real world. Furthermore, we demonstrate that policies trained on data collected in SIMPLE can be transferred zero-shot to physical humanoid robots under similar settings, providing a robust and reproducible foundation for humanoid robotics research.

作者：Songlin Wei, Zhenhao Ni, Jie Liu, Zhenyu Zhao, Junjie Ye, Hongyi Jing, Junkai Xia, Xiawei Liu, Michael Leong, Liang Heng, Di Huang, Yue Wang

来源

本页事实对应的公开页面，以原文为准。

  1. SIMPLE: A Benchmark for Humanoid Loco-Manipulation：SIMPLE: Simulation-Based Policy Learning and Evaluation for Humanoid Loco-manipulation访问 2026-10-10