Toward Global Intent Inference for Human Motion by Inverse Reinforcement Learning

2026-03-08

AI Summary

This study investigates whether a unified cost function can account for and predict human reaching movements without relying on subject- or posture-specific optimization criteria. To this end, the authors propose the Minimum-Observation Inverse Reinforcement Learning (MO-IRL) algorithm, which efficiently estimates time-varying weights from a seven-dimensional set of candidate cost terms to reconstruct planar reaching trajectories. This approach yields, for the first time, a time-varying unified cost function generalizable across individuals and postures, substantially improving prediction accuracy. MO-IRL converges orders of magnitude faster than conventional bilevel optimization methods and requires minimal data. Experimental results demonstrate that incorporating time-varying weights reduces trajectory reconstruction error by 27% on average across all generalization levels, with joint acceleration regularization as the dominant component, complemented by smoothness in torque rate, supporting a unified optimality principle in human motor control.

Technology Category

• Intelligent Robots: Learning & Optimization for ROB
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

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.

Problem

• human motion
• inverse reinforcement learning
• unified cost function
• reaching movements
• optimality principle

Innovation

• Inverse Reinforcement Learning
• Time-varying cost function
• Unified optimality principle
• MO-IRL
• Human motor control