This work proposes a latent flow-based model equipped with a data-driven Gaussian mixture prior that more effectively disentangles diverse human behaviors than conventional single-modal priors and enables natural uncertainty quantification through tractable likelihood computation.
Abstract
Stochastic human motion prediction aims to forecast future motion distributions. Although recent studies have achieved strong performance in terms of accuracy and diversity, they often overlook plausibility (e.g., resulting in physically unrealistic predictions) and uncertainty quantification, both of which are essential for real-world applications and downstream tasks. To address these issues, we propose a latent flow-based model equipped with a data-driven Gaussian mixture prior that more effectively disentangles diverse human behaviors than conventional single-modal priors. This prior is derived from patterns in the training data without requiring additional annotations. Furthermore, the fully invertible nature of our model enables natural uncertainty quantification through tractable likelihood computation. Experiments on the Human3.6M and AMASS datasets demonstrate that our approach achieves state-of-the-art performance in both accuracy and plausibility.
Accurate prediction of human motion using pose information for predicting human trajectories is essential in applications in collaborative environments involving human–robot or human-machine interaction, enabling proactive collision avoidance and a safer workplace. However, forecasting realistic human motion remains...
Deepak Antony David, Luke Busse, Manish Kumar· ASME Letters in Dynamic Syst...· 0 citations
Accurate human trajectory prediction is essential for autonomous driving and robot navigation. Despite substantial progress, deep learning-based approaches often suffer from the train-inference gap caused by distributional discrepancies between training and test environments. The goal-guided framework mitigates this is...
Ge Sun, Jun Ma· IEEE Robotics and Automation...· 0 citations
Object-Conditioned Social Diffusion is proposed, a conditional diffusion model that integrates motion history, multi-person interactions, and object cues into a single framework that reduces the two-second path error, produces more realistic long-term forecasts, and supports sampling multiple plausible futures.
Serdar Ozsoy, Lars Doorenbos, Juergen Gall· 0 citations
This paper presents a self-supervised representation learning framework for understanding 3D skeleton-based human motion in soccer, using future motion prediction as the learning objective. Since human motion is inherently uncertain, accounting for multiple plausible futures is essential for capturing the underlying mo...
Yi Tian Xu, Lars Bretzner, Tiesheng Wang et al.· 0 citations
A unified spatiotemporally decoupled framework named DeMoDiff is proposed, which jointly redesigns representation and architecture and incorporates spatial-temporal masking and attention mechanisms into an autoregressive diffusion generator, achieving both generative capability and controllable editability.
Cheng-Qun Yang, Liang Xu, Yan-Ping Li et al.· 0 citations
GaussianWAM is proposed, a training-time representation-enhancement framework that organizes geometric and semantic supervision through a 3D Gaussian field and improves performance on standard LIBERO and shows positive transfer trends on RoboTwin and real-world manipulation.
Zi-Jian Zhang, Yu-Qing Jiang, Wei-Tao Zhou et al.· 3 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
