Toward Global Intent Inference for Human Motion by Inverse Reinforcement Learning

TL;DR

This study demonstrates that a single, time-varying cost function can accurately predict human reaching movements across different subjects and postures, suggesting a unified optimality principle in human motion.

Contribution

It introduces MO-IRL, a fast inverse reinforcement learning method that estimates time-varying cost weights, enabling practical modeling of human reaching behavior with a unified cost function.

Findings

Time-varying weights improve trajectory prediction by 27% RMSE reduction.

Joint-acceleration regulation is the dominant cost component.

A single cost function predicts trajectories across subjects and postures.

Abstract

This paper investigates whether a single, unified cost function can explain and predict human reaching movements, in contrast with existing approaches that rely on subject- or posture-specific optimization criteria. Using the Minimal Observation Inverse Reinforcement Learning (MO-IRL) algorithm, together with a seven-dimensional set of candidate cost terms, we efficiently estimate time-varying cost weights for a standard planar reaching task. MO-IRL provides orders-of-magnitude faster convergence than bilevel formulations, while using only a fraction of the available data, enabling the practical exploration of time-varying cost structures. Three levels of generality are evaluated: Subject-Dependent Posture-Dependent, Subject-Dependent Posture-Independent, and Subject-Independent Posture-Independent. Across all cases, time-varying weights substantially improve trajectory reconstruction, yielding an average 27% reduction in RMSE compared to the baseline. The inferred costs consistently highlight a dominant role for joint-acceleration regulation, complemented by smaller contributions from torque-change smoothness. Overall, a single subject- and posture-agnostic time-varying cost function is shown to predict human reaching trajectories with high accuracy, supporting the existence of a unified optimality principle governing this class of movements.