MultiPhys: Multi-Person Physics-Aware 3D Motion Estimation
Nicolas Ugrinovic, Boxiao Pan, Georgios Pavlakos, Despoina Paschalidou, Bokui Shen, Jordi Sanchez-Riera, Francesc Moreno-Noguer, Leonidas J. Guibas
摘要
We introduce MultiPhys, a method designed for recovering multi-person motion from monocular videos. Our focus lies in capturing coherent spatial placement between pairs of individuals across varying degrees of engagement. MultiPhys, being physically aware, exhibits robustness to jittering and occlusions, and effectively eliminates penetration issues between the two individuals. We devise a pipeline in which the motion estimated by a kinematic-based method is fed into a physics simulator in an autoregressive manner. We introduce distinct components that enable our model to harness the simulator's properties without compromising the accuracy of the kinematic estimates. This results in final motion estimates that are both kinematically coherent and physically compliant. Extensive evaluations on three challenging datasets characterized by substantial inter-person interaction show that our method significantly reduces errors associated with penetration and foot skating, while performing competitively with the state-of-the-art on motion accuracy and smoothness. Results and code can be found in our project page.
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了每一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。
引用它的顶会 Paper7
- RAM: Recover Any 3D Human Motion in-the-WildSen Jia, Ning Zhu, Jinqin Zhong, Jiale Zhou 等CVPR 2026 · 被引用 12 次
- Ponimator: Unfolding Interactive Pose for Versatile Human-Human Interaction AnimationShaowei Liu, Chuan Guo, Bing Zhou, Jian WangICCV 2025 · 被引用 2 次
- PP-Motion: Physical-Perceptual Fidelity Evaluation for Human Motion GenerationSihan Zhao, Zixuan Wang, Tianyu Luan, Jia Jia 等ACM MM 2025 · 被引用 1 次
- Group Inertial Poser: Multi-Person Pose and Global Translationfrom Sparse Inertial Sensors and Ultra-Wideband RangingYing Xue, Jiaxi Jiang, Rayan Armani, Dominik Hollidt 等ICCV 2025 · 被引用 1 次
- FineTec: Fine-Grained Action Recognition Under Temporal Corruption via Skeleton Decomposition and Sequence CompletionDian Shao, Mingfei Shi, Like LiuAAAI 2026
它引用的顶会 Paper25
- AMASS: Archive of Motion Capture As Surface ShapesNaureen Mahmood, Nima Ghorbani, Nikolaus F. Troje, Gerard Pons-Moll 等ICCV 2019 · 被引用 1,784 次
- HuMoR: 3D Human Motion Model for Robust Pose EstimationDavis Rempe, Tolga Birdal, Aaron Hertzmann, Jimei Yang 等ICCV 2021 · 被引用 398 次
- Monocular, One-stage, Regression of Multiple 3D PeopleYu Sun, Qian Bao, Wu Liu, Yili Fu 等ICCV 2021 · 被引用 335 次
- Putting People in their Place: Monocular Regression of 3D People in DepthYu Sun, Wu Liu, Qian Bao, Yili Fu 等CVPR 2022 · 被引用 152 次
- Dynamics-regulated kinematic policy for egocentric pose estimationZhengyi Luo, Ryo Hachiuma, Ye Yuan, Kris KitaniNeurIPS 2021 · 被引用 130 次
相关 Paper
- Recovering Physically Plausible Human-Object Interactions from Monocular VideosDingbang Huang, Etienne Vouga, Qixing Huang, Georgios PavlakosCVPR 2026
- Closely Interactive Human Reconstruction with Proxemics and Physics-Guided AdaptionBuzhen Huang, Chen Li, Chongyang Xu, Liang Pan 等CVPR 2024
- Trajectory Optimization for Physics-Based Reconstruction of 3d Human Pose from Monocular VideoErik Gärtner, Mykhaylo Andriluka, Hongyi Xu, Cristian SminchisescuCVPR 2022 · 被引用 31 次
- Differentiable Dynamics for Articulated 3d Human Motion ReconstructionErik Gärtner, Mykhaylo Andriluka, Erwin Coumans, Cristian SminchisescuCVPR 2022 · 被引用 33 次
- IF-Garments: Reconstructing Your Intersection-Free Multi-Layered Garments from Monocular VideosMingyang Sun, Qipeng Yan, Zhuoer Liang, Dongliang Kou 等ACM MM 2024 · 被引用 3 次
