3D Reconstruction of Deformable Revolving Object under Heavy Hand Interaction

We reconstruct 3D deformable object through time, in the context of a live making process where the crafter molds the object. Because the object from heavy hand interaction, and is being deformed, classical cannot be applied. We use particle energy optimization to estimate object profile and benefit of the object radial symmetry to increase the of the reconstruction to both occlusion and noise. Our method works an unconstrained scalable setup with one or more depth sensors. We on our database (released upon publication) on a per-frame and basis and shows it significantly outperforms state-of-the-art 7.60mm average object reconstruction error. Further ablation studies the effectiveness of our method.