Temporally-Consistent Surface Reconstruction

In collaboration with colleagues at Adobe Research, we have proposed a method for the unsupervised reconstruction of a temporally-consistent?sequence of surfaces from a sequence of time-evolving point clouds. It yields dense, semantically meaningful correspondences between all keyframes. We represent the reconstructed surface as an atlas that defines canonical correspondences. We leverage them to encourage the reconstruction to be as isometric as possible across frames, leading to semantically-meaningful reconstruction. Through experiments and comparisons, we empirically show that our method achieves results that exceed that state of the art in the accuracy of unsupervised correspondences and accuracy of surface reconstruction.

For more details see our paper that will be presented at ICCV along with this video.