用高斯混合先验提升3D人体运动预测的合理性和不确定性建模
Gaussian-Mixture Latent Flow for Stochastic 3D Human Motion Prediction
- 基于潜空间流模型，引入数据驱动的高斯混合先验以分离多样行为
- 在Human3.6M和AMASS上实现最先进准确率与合理性，且可自然量化不确定性
- 适合需要真实可信轨迹生成的场景，如机器人控制或虚拟人动画
随机人体运动预测旨在生成未来运动的概率分布。尽管近期方法在准确性和多样性上表现优异，却常忽略合理性（如生成物理上不合理的动作）和不确定性量化，而这对于实际应用和下游任务至关重要。为此，我们提出一种基于潜空间流的模型，采用数据驱动的高斯混合先验，相比传统单模态先验更有效地解耦多样人体行为模式，该先验来自训练数据本身，无需额外标注。此外，模型全可逆特性支持通过可计算似然实现自然的不确定性量化。在Human3.6M和AMASS数据集上的实验表明，该方法在准确性和合理性方面均达到当前最优水平。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。