RDT2：探索UMI数据的缩放极限以实现零样本跨具身泛化
RDT2: Exploring the Scaling Limit of UMI Data Towards Zero-Shot Cross-Embodiment Generalization
AI总结：
RDT2通过三阶段训练配方实现零样本跨具身泛化，提升机器人在开放词汇任务中的表现
摘要：
视觉-语言-动作（VLA）模型在通用机器人领域具有潜力，但目前在数据稀缺、架构低效以及无法跨不同硬件平台泛化方面存在困难。我们介绍了RDT2，一个基于70亿参数VLM的机器人基础模型，旨在使模型能够零样本部署到新的具身上以处理开放词汇任务。为此，我们使用增强的、具身无关的通用操作接口（UMI）收集了最大的开源机器人数据集之一——超过10000小时的演示，涵盖多样化的家庭。我们的方法采用了一种新的三阶段训练配方，通过残差向量量化（RVQ）、流匹配和蒸馏将离散语言知识与连续控制对齐，以实现实时推理。因此，RDT2成为首个能够零样本泛化到未见过的对象、场景、指令甚至机器人平台的模型之一。此外，它在灵巧、长周期和动态的下游任务如打桌球中优于最先进的基线。见https://rdt-robotics.github.io/rdt2/了解更多。
英文摘要：
Vision-Language-Action (VLA) models hold promise for generalist robotics but currently struggle with data scarcity, architectural inefficiencies, and the inability to generalize across different hardware platforms. We introduce RDT2, a robotic foundation model built upon a 7B parameter VLM designed to enable zero-shot deployment on novel embodiments for open-vocabulary tasks. To achieve this, we collected one of the largest open-source robotic datasets--over 10,000 hours of demonstrations in diverse families--using an enhanced, embodiment-agnostic Universal Manipulation Interface (UMI). Our approach employs a novel three-stage training recipe that aligns discrete linguistic knowledge with continuous control via Residual Vector Quantization (RVQ), flow-matching, and distillation for real-time inference. Consequently, RDT2 becomes one of the first models that simultaneously zero-shot generalizes to unseen objects, scenes, instructions, and even robotic platforms. Besides, it outperforms state-of-the-art baselines in dexterous, long-horizon, and dynamic downstream tasks like playing table tennis. See https://rdt-robotics.github.io/rdt2/ for more information.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
