Learning Canonical Embeddings for Unsupervised Shape Correspondence with Locally Linear Transformations. (arXiv:2209.02152v2 [cs.CV] UPDATED)

We present a new approach to unsupervised shape correspondence learning
between pairs of point clouds. We make the first attempt to adapt the classical
locally linear embedding algorithm (LLE) -- originally designed for nonlinear
dimensionality reduction -- for shape correspondence. The key idea is to find
dense correspondences between shapes by first obtaining high-dimensional
neighborhood-preserving embeddings of low-dimensional point clouds and
subsequently aligning the source and target embeddings using locally linear
transformations. We demonstrate that learning the embedding using a …

arxiv canonical linear unsupervised