Reconstructing the dynamic motions of humans is one of the most challenging reconstruction problems in computer vision. We develop methods to track individual people in a crowd, detect actions in group activities, and reconstruct the 3D human pose consistently across long videos.
- 3D Human Pose Estimation from Videos
- Tracking Multiple People in a Multi-Camera Environment
- Ball Tracking
- Tracking Interacting Objects
- Learning Monocular Reconstruction from Multi-view Images
- Motion Patterns for Multiple Object Tracking
- Training Procedure for Multiple Object Tracking
- Neural Scene Decomposition