NeurIPS 2020

Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive Object Re-ID


Meta Review

Three of the four reviewers originally recommended marginal accept or accept (7, 6, 6) as they felt the paper provided a good empirical contribution to the field of adaptive re-identification and its results were strong. R9 was more negative and had concerns around experiments. One reviewer pointed out that the DukeMTMC extensively used in the paper has been taken down 12 months ago and its use should be discontinued. Because of the ethical concerns around this, the paper underwent additional review by the ethics panel, which recommended that the dataset should NOT be used in an accepted NeurIPS paper. Some excerpts from the ethics reviewers are below: -- "... the dataset collection involved non-consensual video surveillance of students on Duke University campus. It is unlikely that all students even knew they were being recorded, and their relative lack of power with respect to the institution surveilling them also raises concerns about the ability to meaningfully object to the surveillance." -- "Including the dataset in the paper as-is would be problematic, as it would contribute to this mainstream use of the dataset. Referencing the issues and discouraging future use of the dataset would help mitigate this, as would full removal of the results." -- "The fact that others use the dataset uncritically does not make it appropriate, but it could have contributed to the authors being unaware of the issue, and that awareness may vary geographically." -- "The Broader Impact section should state more clearly that surveillance is the typical goal of re-id systems. It should also state the re-id systems often rely on non-consensual surveillance data for their training. The Broader section should also state that the demographic makeup of the datasets used are not representative of the broader population." -- "...this innovation ... can potential equip malicious actors with the ability of actors to surveil person(s) or groups through multiple cctv cameras without their consent. For potential mitigations, I would strongly urge the authors at a minimum to discuss the potential harms of person re-id due to surveillance, explain the consent challenges with MTMC data collection, as well as remove the Duke MTMC dataset from their evaluation." As a result of the above ethics reviews, the initially positive reviewers have downgraded to marginally below accept, citing concerns around the use of the dataset. The AC read the reviews, rebuttal and ethics reviews and agrees that the paper should remove all experiments on the Duke dataset. However the AC does not think it is fair to reject the paper based on its use of the dataset alone, as its takedown is relatively recent and the authors pointed out that it appears in CVPR 2020 papers, necessitating a comparison. The authors have promised to take the dataset out of the camera ready and replace with experiments on other datasets, where they also promise that similarly strong results can be achieved. The AC therefore recommends a **conditional accept** (a rare and exceptional case reserved only for papers with ethical review), which means the paper is accepted on the condition that the final version 1) removes all results on DukeMTMC, 2) makes it clear that the dataset has been taken down and should no longer be used, and 3) reproduces all technical contributions and results of the original submission using other datasets. The authors are also encouraged to take into account other reviewer suggestions in preparing the camera ready. The AC discussed this decision with the SAC and PCs. ******************************* Note from Program Chairs: The camera-ready version of this paper has been reviewed with regard to the conditions listed above, and this paper is now fully accepted for publication.