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

TL;DR

This paper introduces ProbHMI, a novel invertible network-based approach for 3D human motion forecasting that explicitly models uncertainty, improving safety and reliability in autonomous systems.

Contribution

The paper presents ProbHMI, an invertible network framework that disentangles pose representations and explicitly predicts future uncertainty in 3D human motion forecasting.

Findings

ProbHMI achieves state-of-the-art accuracy in deterministic and diverse motion prediction.

The model effectively calibrates uncertainty estimates, enhancing risk-aware decision making.

ProbHMI outperforms existing methods on benchmark datasets.

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.