A Spatio-temporal Continuous Network for Stochastic 3D Human Motion Prediction

2025-08-03

AI Summary

Existing methods for 3D human motion prediction struggle to jointly model stochasticity and continuous-time dynamics, often suffering from mode collapse, limited diversity, and kinematically implausible motion trajectories. To address these issues, we propose the Spatio-Temporal Continuous Network (STCN), which introduces an anchor set mechanism to explicitly represent latent motion modalities and mitigate mode collapse. STCN further employs a Gaussian Mixture Model (GMM) to perform probabilistic modeling of observed motion sequences and enhances generative diversity via multi-sequence sampling. Evaluated on Human3.6M and HumanEva-I, STCN achieves significant improvements in both prediction accuracy and motion diversity. The generated motions exhibit enhanced physical plausibility and statistical robustness. By unifying stochastic modeling with continuous-time dynamics, STCN establishes a scalable, continuous probabilistic framework for random human motion generation.

Technology Category

• Reasoning under Uncertainty: Relational Probabilistic Models
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Computer Vision: Motion & Tracking

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Stochastic Human Motion Prediction (HMP) has received increasing attention due to its wide applications. Despite the rapid progress in generative fields, existing methods often face challenges in learning continuous temporal dynamics and predicting stochastic motion sequences. They tend to overlook the flexibility inherent in complex human motions and are prone to mode collapse. To alleviate these issues, we propose a novel method called STCN, for stochastic and continuous human motion prediction, which consists of two stages. Specifically, in the first stage, we propose a spatio-temporal continuous network to generate smoother human motion sequences. In addition, the anchor set is innovatively introduced into the stochastic HMP task to prevent mode collapse, which refers to the potential human motion patterns. In the second stage, STCN endeavors to acquire the Gaussian mixture distribution (GMM) of observed motion sequences with the aid of the anchor set. It also focuses on the probability associated with each anchor, and employs the strategy of sampling multiple sequences from each anchor to alleviate intra-class differences in human motions. Experimental results on two widely-used datasets (Human3.6M and HumanEva-I) demonstrate that our model obtains competitive performance on both diversity and accuracy.

Problem

• Learning continuous temporal dynamics in human motion prediction
• Preventing mode collapse in stochastic motion sequences
• Reducing intra-class differences in complex human motions

Innovation

• Spatio-temporal continuous network for smoother motion
• Anchor set prevents mode collapse in predictions
• GMM and multi-sequence sampling reduce intra-class differences