NeurIPS 2020

Neural Star Domain as Primitive Representation

Meta Review

This work introduces a new representation for 3D shapes based on primitives defined on a starred domain. Reviewers find the proposed approach novel and the empirical results convincing. Reviewers brought up some weaknesses mostly related to clarity, which were sufficiently addressed by the rebuttal. All reviewers recommend accept after the rebuttal. AC did not find sufficient grounds to overturn this consensus recommendation. The authors should revise the paper by incorporating the discussion in the rebuttal.