I thank the authors for their submission. The paper presents an algorithm for unsupervised object segmentation. I strongly encourage the authors to take into account the reviewers' comments and concerns for the final manuscript, in particular regarding failure points, weaknesses and directions for future work.