Review History


All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.

Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.

View examples of open peer review.

Summary

  • The initial submission of this article was received on November 11th, 2020 and was peer-reviewed by 2 reviewers and the Academic Editor.
  • The Academic Editor made their initial decision on February 12th, 2021.
  • The first revision was submitted on February 28th, 2021 and was reviewed by the Academic Editor.
  • The article was Accepted by the Academic Editor on March 1st, 2021.

Version 0.2 (accepted)

· Mar 1, 2021 · Academic Editor

Accept

Thanks for the revision. The paper is ready to be accepted.

Version 0.1 (original submission)

· Feb 12, 2021 · Academic Editor

Minor Revisions

Both reviewers would like the authors to improve the paper and clarify the concerns raised.

·

Basic reporting

The writing of this article is particularly good to get what readers are concerned about. Thus, a positive recommendation would be provided under good expectations.

Experimental design

The authors divide the images using fixed sizes to create two databases. The partition of the database should reference the popular way in the deep learning community as well. For example, creating each corresponding steganographic image for each cover image and structuring it as a complete database, then allowing the designed algorithm to select appropriate proportions. Besides, giving a clear name to illustrate the combination of BOSSBase V1.01 and BOWS2 databases. Better to make a website based on GitHub to introduce the summarized databases.

Validity of the findings

The presented innovation meets the requirements that obtain a pleasant recommendation. Better to convert the current study to a benchmark for other researchers. Also, the strategy enclosing special elements might lack a scientific basis. Most of them might be empirical and there are difficulties to learn systematically. Of course, true actions depend on the authors.

Additional comments

Dear authors, so glad to read your paper regarding investigating the CNN (convolutional neural network) based methods applied to steganalysis and adding special modules to enhance the resultant accuracy.

Reviewer 2 ·

Basic reporting

no comment

Experimental design

no comment

Validity of the findings

Novel enough for publication.

Additional comments

The authors propose a general strategy to improve existing deep learning based steganalytic methods. The strategy makes modifications in the stages of pre-processing, feature extraction, and classification using SRM filters, batch normalization, absolute value layer, and spatial dropout. The proposed scheme is novel and interesting. The manuscript is suitable for publication. Some comments are as follows:
- The proposed scheme makes modifications in several stages of deep steganalysis. The effectiveness of modifications in each stage should be discussed. For example, some experimental results should be given.
- It is better to summarize the contributions of the manuscript in the end of Introduction.

All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.