Submissions

Reviewer Guidelines

Reviewer Guidelines

Thank you for volunteering to review for ACCV 2024!  Our community greatly relies on your time and expertise to ensure a high-quality program.  The following outlines expectations and policies for reviewers:

  • Review deadline: all reviews are due on August 12, 2024.  If you are unable to complete your reviews by the deadline, please contact your area chair as soon as possible so that a replacement or emergency reviewer can be assigned.
  • Assignment check: As soon as the reviewing assignments are given, please go through them to check for (a) no obvious conflicts, perhaps due to overlooked declarations (b) a reasonable match to your area of expertise and (c) obvious desk rejects.  If issues arise, please notify your area chair immediately.
  • Familiarization with author guidelines: please read the Author Guidelines regarding policies on anonymization, dual submissions, plagiarism, etc. If you suspect any violations, please notify your area chair immediately.  In the meantime, please proceed to review the paper.
  • Anonymization & Double blind reviews: ACCV’s review process is double blind.  The authors do not know the identity of the reviewers or area chairs, while the area chairs and reviewers should not infer the author identities. Authors are instructed to anonymize their submissions by not disclosing their identity, institutions and affiliations, acknowledgments and funding sources.  Similarly, reviewers should not disclose any identifying information in their reviews.
  • ArXiv Preprints: Submissions are sometimes already released on arXiv.  This does not violate dual submission policies.  Please do not attempt to find out the authors by searching for pre-prints.  Do your best to treat your assignments impartially, whether or not you know, or suspect, the author identities. Reviewers may suggest related works on arXiv, but should not reject papers solely for the lack of discussion, comparison or citation of arXiv preprints, nor on grounds of similar ideas to arXiv preprints.  Instead, any suspicion of plagiarism or dishonesty should be brought up separately to the area chairs.
  • LLM usage: Reviews featuring text generated from a large-scale language model (LLM) such as ChatGPT are strictly prohibited.  Submission content is expected to stay confidential and should therefore not be passed to online platforms like ChatGPT.
  • Good reviewing practice: Please read the ACCV 2024 Good Practices in Review tutorial.  Look out for papers that are technically sound and make contributions to the field. Please embrace new concepts, even if they have not been tested on many datasets.  Please weigh both novelty and potential impact of a submission along with its reported performance; results which do not exceed state-of-the-art accuracy on existing benchmarks are not sole grounds for rejection.  Similarly, minor flaws which are easily correctable should also not be grounds for rejection.
  • Other checks:
    • Reproducibility: Authors are highly encouraged (but not required) to submit their source code as a part of the supplementary material.  Reviewers may (but are not required to) check the code if they wish to ensure reproducibility.  Any code should be reviewed confidentially and deleted after the review process.
    • Data contribution: papers that claim data contribution are expected to release their data publicly by camera-ready deadline.  If a submission claims a dataset contribution, please confirm this in the reviewing checklist.
      Attribution and citations: Please check if a paper has adequately cited papers, data and code assets used in the paper and comment accordingly.  Missing attribution should not be the sole ground for rejection.  If academic dishonesty is suspected, please notify the ACs.
    • Discussion on limitations: Authors are highly encouraged to discuss limitations of their work in a forthcoming manner.  Reviewers should consider such a discussion in a positive light and not penalize authors, nor treat the limitations directly as weaknesses of the work.
    • Discussion on negative societal impacts: Authors are encouraged to include a discussion on negative societal impacts of their submission.  A lack of such a discussion should not be the sole grounds for rejection.
    • Personal & Human Subject Data: Please familiarize yourself with the Ethics Guidelines.  Papers which collect or use personal data or data from human subjects should either already have IRB or equivalent body approval; otherwise it should clearly describe the ethical principles that have been followed.  If such a description is missing, or there are glaring violations of the ethics guidelines, please inform your AC immediately.