PhysPT: Physics-aware Pretrained Transformer for Estimating Human Dynamics from Monocular Videos
CVPR 2024（2024）
关键词
Monocular Video,Training Loss,Contact Force,Physical Principles,Human Motion,Loss Of Contact,Motion Estimation,3D Motion,Motion Force,Human Activity Recognition,Loss Of Force,Human 3D,Self-supervised Manner,Deep Models,Body Movements,3D Reconstruction,Body Shape,Motion Capture,Reconstruction Accuracy,Force Estimation,3D Body,Lagrange Equations,Physics Engine,3D Human Model,Self-supervised Learning,Global Trajectory,Body Pose,3D Motion Capture,Contact Model,Left Foot
AI 理解论文
溯源树
样例
生成溯源树，研究论文发展脉络
Chat Paper
正在生成论文摘要