Planning safe robot motions in the presence of humans requires reliable forecasts of future human motion. However, simply predicting the most likely motion from prior interactions does not guarantee safety. Such forecasts fail to model the long tail of possible events, which are rarely observed in limited datasets. On the other hand, planning for worst-case motions leads to overtly conservative behavior and a ``frozen robot''. Instead, we aim to learn forecasts that predict counterfactuals that humans guard against. We propose a novel game-theoretic framework for joint planning and forecasting with the payoff being the performance of the planner against the demonstrator, and present practical algorithms to train models in an end-to-end fashion. We demonstrate that our proposed algorithm results in safer plans in a crowd navigation simulator and real-world datasets of pedestrian motion. We release our code at https://github.com/portal-cornell/Game-Theoretic-Forecasting-Planning.

翻译：在人类环境中规划安全的机器人运动需要可靠的人类未来运动预测。然而，仅根据先验交互预测最可能的运动并不能保证安全性。此类预测未能建模罕见事件的长尾分布，这些事件在有限数据集中难以被观测。另一方面，针对最坏情况运动进行规划则会导致过度保守的行为和“机器人冻结”现象。为此，我们旨在学习能够预测人类主动防范的反事实路径的模型。本文提出一种新颖的博弈论联合规划与预测框架，将规划器相对于演示者的性能作为收益函数，并给出端到端训练模型的实际算法。我们在人群导航模拟器和真实世界行人运动数据集上证明，所提算法能够生成更安全的规划方案。代码开源地址：https://github.com/portal-cornell/Game-Theoretic-Forecasting-Planning。