21 dynamic scenes captured by 53 synchronized cameras, providing high-resolution multi-view video and 3D reconstructions. The dataset includes diverse human activities and object interactions recorded from a 360-degree perspective for volumetric video research.
Use Cases
- Train neural radiance fields (NeRF) or 3D Gaussian Splatting models using the synchronized multi-view video frames and camera poses
- Evaluate dynamic scene reconstruction algorithms using the provided 3D point cloud sequences as ground truth
- Develop human pose estimation models that leverage 53 simultaneous viewpoints for occlusion-resistant tracking
Strengths
- 53 synchronized camera views providing full 360-degree coverage of dynamic scenes
- 21 distinct dynamic sequences featuring human-object interactions and complex motions
- Includes calibrated camera parameters, foreground masks, and reconstructed 3D point clouds