轻量化建模多人运动交互，高效预测且精度领先。
Efficient Multi-Person Motion Prediction by Lightweight Spatial and Temporal Interactions
- 双分支结构分别学习个体与群体的局部和全局表征。
- 在CMU-Mocap等3个数据集上达到顶尖性能，计算量显著降低。
- 引入距离嵌入增强空间交互建模，适合实时多人体运动预测场景。
3D多人运动预测是一项高度复杂的任务，主要源于个体历史运动与人之间交互的依赖性。有效建模这些交互通常带来巨大的计算开销。本文提出一种计算高效的多人运动预测模型，通过简化空间与时间交互来实现。方法设计轻量级双分支结构，分别学习单人与多人的局部和全局表征；引入新型跨层级交互模块，融合双分支的空间与时间特征；并显式加入人之间的空间距离嵌入以增强交互建模。该设计在标准数据集CMU-Mocap、MuPoTS-3D和3DPW上实现多项指标领先，同时大幅降低计算成本。代码已开源：https://github.com/Yuanhong-Zheng/EMPMP。
原文摘要 · Abstract (English)
3D multi-person motion prediction is a highly complex task, primarily due to the dependencies on both individual past movements and the interactions between agents. Moreover, effectively modeling these interactions often incurs substantial computational costs. In this work, we propose a computationally efficient model for multi-person motion prediction by simplifying spatial and temporal interactions. Our approach begins with the design of lightweight dual branches that learn local and global representations for individual and multiple persons separately. Additionally, we introduce a novel cross-level interaction block to integrate the spatial and temporal representations from both branches. To further enhance interaction modeling, we explicitly incorporate the spatial inter-person distance embedding. With above efficient temporal and spatial design, we achieve state-of-the-art performance for multiple metrics on standard datasets of CMU-Mocap, MuPoTS-3D, and 3DPW, while significantly reducing the computational cost. Code is available at https://github.com/Yuanhong-Zheng/EMPMP.
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。