Uncertainty-aware Probabilistic 3D Human Motion Forecasting via Invertible Networks

2025-07-19

AI Summary

In safety-critical human–machine interaction (HMI) scenarios, quantifying uncertainty in 3D human motion prediction remains challenging due to poor calibration and uncontrolled risk arising from implicit probabilistic modeling in existing approaches. To address this, we propose ProbHMI—a novel framework leveraging invertible neural networks to construct a disentangled latent space, where probabilistic dynamics are explicitly modeled to jointly optimize pose sequence generation and uncertainty propagation. ProbHMI is the first method to simultaneously support diverse trajectory sampling and strictly calibrated uncertainty estimation, enabling quantile-level confidence outputs and risk-aware decision-making. Extensive evaluation on multiple benchmarks demonstrates that ProbHMI achieves state-of-the-art performance in deterministic accuracy, predictive diversity, and uncertainty calibration. By providing reliable, well-calibrated probabilistic forecasts, ProbHMI establishes a robust foundation for safety-sensitive human motion prediction in HMI systems.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Probabilistic Programming
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

3D human motion forecasting aims to enable autonomous applications. Estimating uncertainty for each prediction (i.e., confidence based on probability density or quantile) is essential for safety-critical contexts like human-robot collaboration to minimize risks. However, existing diverse motion forecasting approaches struggle with uncertainty quantification due to implicit probabilistic representations hindering uncertainty modeling. We propose ProbHMI, which introduces invertible networks to parameterize poses in a disentangled latent space, enabling probabilistic dynamics modeling. A forecasting module then explicitly predicts future latent distributions, allowing effective uncertainty quantification. Evaluated on benchmarks, ProbHMI achieves strong performance for both deterministic and diverse prediction while validating uncertainty calibration, critical for risk-aware decision making.

Problem

• Quantify uncertainty in 3D human motion forecasting
• Enable probabilistic modeling via invertible networks
• Improve risk-aware decision-making in human-robot collaboration

Innovation

• Invertible networks for disentangled latent space
• Explicit future latent distributions prediction
• Effective uncertainty quantification and calibration