随笔分类 -  6 具身智能

[思考] 具身智能
摘要：目录人形机器人基于动力学方法盲人行走法传感器+世界模型更多动作的训练？更难动作的训练？VLA模型数据Open X-Embodiment (OXE)仿真环境数据集大规模人类视频数据集评测核心指标：任务成功率 (Success Rate, SR)评测场景：仿真 vs. 真实世界泛化能力基准测试 (Gen 阅读全文
posted @ 2025-08-20 17:50 fariver 阅读(83) 评论(0) 推荐(0)
[PaperReading] π0.5: a Vision-Language-Action Model with Open-World Generalization
摘要：目录π0.5: a Vision-Language-Action Model with Open-World GeneralizationTL;DRMethodTraining RecipeData硬件配置Experiment效果可视化总结与思考相关链接 π0.5: a Vision-Languag 阅读全文
posted @ 2025-08-16 17:39 fariver 阅读(79) 评论(0) 推荐(0)
[PaperReading] GR00T N1: An Open Foundation Model for Generalist Humanoid Robots
摘要：目录GR00T N1: An Open Foundation Model for Generalist Humanoid RobotsTL;DRMethodSystem2System1Training SettingDatahuman video数据pretrain方法神经轨迹生成流程仿真轨迹生成方 阅读全文
posted @ 2025-08-14 22:20 fariver 阅读(31) 评论(0) 推荐(0)
[PaperReading] GR-2: A Generative Video-Language-Action Model with Web-Scale Knowledge for Robot Manipulation
摘要：目录GR-2: A Generative Video-Language-Action Model with Web-Scale Knowledge for Robot ManipulationTL;DRMethod模型架构关于cVAEDatasetPre-train DatasetRobotic D 阅读全文
posted @ 2025-08-12 22:11 fariver 阅读(26) 评论(0) 推荐(0)
[PaperReading] RT-1: ROBOTICS TRANSFORMER FOR REAL-WORLD CONTROL AT SCALE
摘要：目录RT-1: ROBOTICS TRANSFORMER FOR REAL-WORLD CONTROL AT SCALETL;DRMethodinstructions与images模态特征处理TokenLearnerAction预测DataExperiment在新任务上Zero-shot泛化能力仿真 阅读全文
posted @ 2025-08-12 20:46 fariver 阅读(45) 评论(0) 推荐(0)
[PaperReading] Helix: A Vision-Language-Action Model for Generalist Humanoid Control
摘要：目录Helix: A Vision-Language-Action Model for Generalist Humanoid ControlTL;DRMethodMotivationSystem 2 (S2，慢系统)System 1 (S1, 快系统)DataExperiment效果可视化总结与思 阅读全文
posted @ 2025-08-11 21:38 fariver 阅读(105) 评论(0) 推荐(0)
[PaperReading] RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control
摘要：目录RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic ControlTL;DRMethod模型Action表示Co-Fine-TuneReal-Time Inference如何实现连续运动控制训练数据Exper 阅读全文
posted @ 2025-08-09 14:33 fariver 阅读(45) 评论(0) 推荐(0)
[PaperReading] π0: A Vision-Language-Action Flow Model for General Robot Control
摘要：目录π0: A Vision-Language-Action Flow Model for General Robot ControlTL;DRMethodPaliGemma VLM基座模型VLA多模态的对齐机制与Transfusion的联系Flow Matching生成如何在本文所介绍的VLA模型 阅读全文
posted @ 2025-08-08 21:46 fariver 阅读(191) 评论(0) 推荐(0)
[PaperReading] GR-1: UNLEASHING LARGE-SCALE VIDEO GENERATIVE PRE-TRAINING FOR VISUAL ROBOT MANIPULATION
摘要：目录UNLEASHING LARGE-SCALE VIDEO GENERATIVE PRE-TRAINING FOR VISUAL ROBOT MANIPULATIONTL;DRMethodPretrainRobot Data FinetuningExperiment总结与思考相关链接 UNLEAS 阅读全文
posted @ 2025-08-07 21:48 fariver 阅读(16) 评论(0) 推荐(0)
[PaperReading] OpenVLA: An Open-Source Vision-Language-Action Model
摘要：目录OpenVLA: An Open-Source Vision-Language-Action ModelTL;DRMethodaction表示Training DataImplementationInfrastructureExperiment效果可视化总结与思考相关链接Related work 阅读全文
posted @ 2025-08-07 20:18 fariver 阅读(152) 评论(0) 推荐(0)
[Paper Reading] HPT: Scaling Proprioceptive-Visual Learning with Heterogeneous Pre-trained Transformers
摘要：目录Scaling Proprioceptive-Visual Learning with Heterogeneous Pre-trained TransformersTL;DRMethodStemTrunkLossHeadExperiment训练资源效果可视化总结与发散相关链接资料查询 Scali 阅读全文
posted @ 2024-10-11 21:11 fariver 阅读(235) 评论(0) 推荐(0)
[Embodied AI Tutorial] The Basic Frameworks and Techniques for Embodied AI (Part2)
摘要：目录Embodied AI Tutorial课程内容Modeling and approaches for Embodied AIWorld ModelGet a Good PolicyPlanning And ControlSimulation technology for Embodied AI 阅读全文
posted @ 2024-09-10 21:48 fariver 阅读(139) 评论(0) 推荐(0)
[Embodied AI Tutorial] Overview of Embodied AI (Part1)
摘要：目录Overview of Embodied AI (Part1)课程内容SimulatorsAssetsTask资料查询 Overview of Embodied AI (Part1) 课程主页：https://ai-workshops.github.io/building-and-working 阅读全文
posted @ 2024-09-10 19:28 fariver 阅读(112) 评论(0) 推荐(0)
[行业调研] 2024世界机器人大会
摘要：展会见闻 从身体结构的维度 机械臂: 吸取小物体，放到另一个区域。机械弊对于这种任务在各种环境下都能精准完成，相对于人类有优势。 灵巧手: 看到各种自由度的手，手掌/手指正面覆盖了很多触觉传感器，有些甚至还布置了摄像头。 AGV: 仓库托运重物，物流机器人。 仿生机器人: 这类公司注重还原外形面部表 阅读全文
posted @ 2024-08-25 17:32 fariver 阅读(71) 评论(0) 推荐(0)