Review guidelines

This document provides essential guidelines for SCIA 2025 reviewers. It outlines what constitutes a quality review and our expectations for your role. Additionally, it explains the anonymization rules for the double-blind peer review process. Please refer to these guidelines as part of the SCIA 2025 review process.

Reviewers will receive formal recognition for their valuable contributions, with their names acknowledged in the conference proceedings.

What Makes a Good Review

As a reviewer for SCIA 2025, your task is to identify outstanding papers that will significantly contribute to the image analysis community. Your expert evaluation helps the program committee select papers of wide interest and potential impact. A good review provides a well-supported assessment of the paper’s strengths and weaknesses.

Key components of your review include:

  1. Summary of the Paper: Provide a brief overview of the paper’s contributions, methods, and results. This helps both the program committee and authors understand the key elements of the work and verify that you’ve captured the essence of the paper.
  2. Assessment of Strengths: Highlight and justify the paper’s key strengths, such as innovative methods, strong evaluations, or unique applications. Explain why these aspects are valuable to the field.
  3. Assessment of Weaknesses: List and justify any major concerns, referencing prior work if the method lacks novelty or pointing out areas that need improvement.
  4. Clarity and Organization: Comment on how clearly the paper is written and organized. Indicate whether it is easy to understand or if improvements are needed in presentation.
  5. Reproducibility: Evaluate whether the paper provides enough detail for others to replicate the work. Note the availability of open data or code and the clarity of methodological descriptions.
  6. Constructive Feedback: Offer constructive criticism to help authors refine their work, including suggestions for minor corrections.
  7. Recommendation: Recommend whether the paper should be accepted, considering its overall contribution to the field. Remember that impactful papers aren’t limited to novel algorithms—they can also present new applications or insights.
  8. Justification of Recommendation: Clearly explain the reasoning behind your recommendation, balancing the strengths and weaknesses.
  9. Ranking: Rank the paper among those you’ve reviewed to assist in calibrating your ratings.
  10. Reviewer Expertise: State your level of expertise in the paper’s subject area, noting any limitations in your understanding.

What to Avoid in a Review

When conducting your review for SCIA 2025, please steer clear of the following:

  1. Surface-Level Summaries: Avoid merely summarizing the paper and focusing on minor details. Your review should engage with the core contributions and provide meaningful feedback on the key aspects of the work.
  2. Unsupported Opinions: Do not express opinions without clear justification. If you find a method novel, specify what makes it novel and why it’s important. Conversely, if you believe it’s not novel, explain why and cite relevant prior work.
  3. Impoliteness: Maintain a polite and respectful tone. Constructive criticism is more effective and fosters a positive dialogue.
  4. Requests for Major Revisions: Refrain from asking authors to make significant expansions to their paper. Evaluate the submission as it stands, as there is no guarantee that suggested changes can be implemented within the scope of the conference.

General Review Considerations

SCIA 2025 encompasses a wide range of topics in image analysis, computer vision, and related fields. To ensure a balanced selection of papers across all relevant topics, please consider the following guidelines during your review process.

When reviewing SCIA papers, focus on:

  • The novelty and innovation of the proposed methods or applications.
  • The relevance of the work to the specific topics within the scope of SCIA 2025.

Key questions to ask include:

  • Does the paper address a significant challenge in its respective topic area?
  • Are the methods clearly described, including data collection, processing, and analysis techniques?
  • Does the data used appropriately represent the range and diversity relevant to the problem being addressed?
  • Are the performance metrics comprehensive, and do they include measures of uncertainty or confidence?
  • How well do the results compare with existing methods, and are they effectively contextualized in terms of significance and contribution to the field?
  • Is the work a substantial contribution to its topic, or is it primarily an incremental improvement over prior research?
  • Do the authors discuss the limitations of their approach and suggest directions for future research?

Your reviews should help ensure that SCIA 2025 presents a diverse and high-quality program that reflects the latest advancements and practical innovations across the full spectrum of computer vision and image analysis topics.

Formal Rules

Confidentiality. As a reviewer for SCIA 2025, you must protect the confidentiality of the papers you review. SCIA submissions are proprietary and unpublished, and submitting them for review does not constitute public disclosure. Therefore, please follow these guidelines:

  • Do not share the paper with anyone not directly involved in the review process. If you involve colleagues or students, they must also maintain confidentiality.
  • Use of LLMs: You may use general writing tools (e.g., LLMs) for grammar and wording but are fully responsible for your review’s content. Do not share any part of the paper with an LLM.
  • Do not disclose supplementary materials (e.g., images, videos) to non-reviewers.
  • Do not use ideas from the paper for your own work before its publication.
  • After reviewing, destroy all copies of the paper and supplementary materials.

Conflict of Interest. The double-blind review process ensures anonymity of authorship. If you suspect you recognize the work or its authors and this may create a conflict of interest, decline the review and notify the Area Chair and Program Chairs. Conflicts of interest include:

  • Current or recent (within three years) affiliation with the same institution.
  • Co-authorship with the author(s) in the past three years.
  • Holding or applying for a grant together in the past three years.
  • Current or planned collaboration.
  • Business relationships or partnerships.
  • Close personal or familial relationships.

Anonymization Rules

SCIA 2025 follows a double-blind review process, where the identities of both reviewers and authors remain anonymous throughout submission, review, and rebuttal.

Ensuring Anonymity:

  • For Authors: Authors must not include names, affiliations, or acknowledgments in their submissions. This information will be added to the final, camera-ready version. Refer to the Submission Guidelines for detailed instructions.
  • For Reviewers: Reviewers must avoid revealing their identity. Do not suggest that authors cite your work unless absolutely necessary, as this could compromise your anonymity.
  • Accidental Discovery: If you accidentally learn the authors’ identity, remain impartial. Do not let this knowledge affect your review. Report any breaches of anonymization to the Program Chairs.

ArXiv Papers:

  • Accidental Identification: Do not attempt to identify authors through arXiv or other public technical reports. If you do discover the authors’ identity by accident, this should not influence your review.
  • Citing ArXiv: ArXiv papers are not considered prior work as they have not been peer-reviewed. Do not penalize submissions for not citing arXiv papers.

Thank you in advance for your dedication and contributions to making SCIA 2025 a successful conference.

SCIA 2025 Program Chairs