In collaborative human-robot manipulation, a robot must predict human intents and adapt its actions accordingly to smoothly execute tasks. However, the human's intent in turn depends on actions the robot takes, creating a chicken-or-egg problem. Prior methods ignore such inter-dependency and instead train marginal intent prediction models independent of robot actions. This is because training conditional models is hard given a lack of paired human-robot interaction datasets. Can we instead leverage large-scale human-human interaction data that is more easily accessible? Our key insight is to exploit a correspondence between human and robot actions that enables transfer learning from human-human to human-robot data. We propose a novel architecture, InteRACT, that pre-trains a conditional intent prediction model on large human-human datasets and fine-tunes on a small human-robot dataset. We evaluate on a set of real-world collaborative human-robot manipulation tasks and show that our conditional model improves over various marginal baselines. We also introduce new techniques to tele-operate a 7-DoF robot arm and collect a diverse range of human-robot collaborative manipulation data, which we open-source.

翻译：在人机协同操作中，机器人需预测人类意图并相应调整自身动作以流畅执行任务。然而，人类意图反过来又取决于机器人采取的动作，这形成了先有鸡还是先有蛋的问题。现有方法忽略了这种相互依赖关系，而是训练与机器人动作无关的边际意图预测模型。其根本原因在于，由于缺乏配对的人机交互数据集，训练条件预测模型十分困难。我们能否转而利用更易获取的大规模人人交互数据？本文的关键洞察在于挖掘人与机器人动作之间的对应关系，从而实现从人人数据到人机数据的迁移学习。我们提出新型架构InteRACT，它在大规模人人数据集上预训练条件意图预测模型，并在小规模人机数据集上进行微调。通过一系列真实世界协同人机操作任务的评估，我们的条件模型较各类边际基线方法均有显著提升。我们还引入了遥操作七自由度机械臂的新技术，采集并开源了多样化的人机协同操作数据。