NeurIPS 2020

Delving into the Cyclic Mechanism in Semi-supervised Video Object Segmentation

Meta Review

After rebuttal and discussion, all four reviewers have arrived to the consensus that the paper should be accepted. The reviewers acknowledge the nice idea and rationale behind the paper. Initially, the experimental results were not entirely convincing, but the extra experiments in the rebuttal showed that the improvements found were due to the novel method. This decision was verified by the AC and SAC. Based on feedback from the reviewers, the authors are requested to - make clear in the experiments the full comparison of their work versus STM and explain any discrepancies - provide more visualizations to help clarify their text and explanations. - add the recommended related work The reviewers also strongly advise that authors also add the experiments using extra data to make the comparisons more meaningful.