NeurIPS 2020

Deep Shells: Unsupervised Shape Correspondence with Optimal Transport

Meta Review

Although there was a substantial amount of discussion, most of the reviewers were excited about the new ideas in your work and their promise for improving shape correspondence. R2 has provided a detailed list of suggested changes for your final revision in their post-rebuttal comments; please address these in revising your paper for the camera-ready. Please also make sure your text clearly articulates the relationship of your work to GeomFmap (it's OK to call it "concurrent work" given the timing), and make all changes/clarifications promised in your rebuttal.