A Game-Theoretic Framework for Joint Forecasting and Planning

08/11/2023
by   Kushal Kedia, et al.
cornell university
0

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.

READ FULL TEXT

page 4

page 5

03/08/2022

Motron: Multimodal Probabilistic Human Motion Forecasting

Autonomous systems and humans are increasingly sharing the same space. R...
09/22/2022

Robust Forecasting for Robotic Control: A Game-Theoretic Approach

Modern robots require accurate forecasts to make optimal decisions in th...
01/07/2021

Safety-Oriented Pedestrian Motion and Scene Occupancy Forecasting

In this paper, we address the important problem in self-driving of forec...
10/04/2022

RAP: Risk-Aware Prediction for Robust Planning

Robust planning in interactive scenarios requires predicting the uncerta...
07/23/2020

Implicit Latent Variable Model for Scene-Consistent Motion Forecasting

In order to plan a safe maneuver an autonomous vehicle must accurately p...
11/16/2020

LUCIDGames: Online Unscented Inverse Dynamic Games for Adaptive Trajectory Prediction and Planning

Existing game-theoretic planning methods assume that the robot knows the...
03/09/2021

On complementing end-to-end human motion predictors with planning

High capacity end-to-end approaches for human motion prediction have the...

Please sign up or login with your details

Forgot password? Click here to reset