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.
All the concerns have been addressed well in the revised version and all the reviewers satisfy the quality of the revision. Therefore, I recommend accepting this manuscript as it is.
[# PeerJ Staff Note - this decision was reviewed and approved by Rong Qu, a PeerJ Computer Science Section Editor covering this Section #]
The authors have considered the comments and suggestions to improve the manuscript. The contributions have been indicated in the revised manuscript. The typos and grammatical errors have been fixed. Some long sentences have been shortened to improve the readability. Vector-based images have been used to improve readability.
The reviewer's comments have been considered and the revised manuscript is satisfactory. The selection of key user-defined parameters has been clearly stated. The missing references are properly cited in the revised manuscript. Further explanation for Figs 7 and 8 have been added to improve the readability.
Rationale and benefit to literature are clearly stated. The conclusions are well stated and linked to the research questions. All underlying data are robust and statistically sound.
The author has made improvements and answered my question.
The author has made improvements and answered my question.
The author has made improvements and answered my question.
The topic is interesting and the contribution is acceptable. However, the novelty of the paper should be further highlighted as the current presentation cannot distinguish the results from the existing ones. The investigated problem should be described formally with more details and the typos need to be corrected before resubmitting the revised version.
[# PeerJ Staff Note: Please ensure that all review and editorial comments are addressed in a response letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate. #]
[# PeerJ Staff Note: The Academic Editor has identified that the English language must be improved. PeerJ can provide language editing services - please contact us at copyediting@peerj.com for pricing (be sure to provide your manuscript number and title) #]
This manuscript studies a simplified Swin Transformer model for oÿine handwritten Chinese character recognition. The authors considered three hierarchical stages and managed to enhance the information interaction between the window and the window. It is a topic of interest in pattern recognition and computer vision. However, the contributions are not clearly indicated, and there are some minor issues that need to be considered and addressed.
1. The original contributions of this paper should be emphasized compared with existing works.
2. There are some typos and grammatical errors in the manuscript. It is strongly suggested that the whole work be carefully checked by someone who has expertise in technical English writing.
3. Some sentences are too long to follow, it is suggested to break them down into short but meaningful ones to make the manuscript readable.
4. Some of the figures are difficult to observe and read. Please use vector-based images to improve the readability.
1. There are many user-defined parameters in the proposed model. How to select and optimize these parameters?
2. The authors compare the performance of their approach with some of the latest approaches, e.g., AlexNet, VGG16, etc. However, these approaches are not properly cited in the paper.
3. In Figure 7 and Figure 8, it is not clear what are the performances of validation accuracy, change of loss function and so on. These should be clearly described and analysed in the experiment results analysis.
4. Parameter selection and rationale of each method implemented should be indicated in the experiment design.
Overall, the authors have presented some interesting results and step forward based on previous works. Conclusions are well stated, and linked to the original research question to support the results. The paper is well organized and logically explained. The quality of the paper can be further improved by addressing the questions and problems raised in this review report.
The article is clearly structured and easy to follow. Comments are listed as follows:
1. What is the different between offline HCCR and online HCCR? Why this work focus on offline HCCR? This should be briefly discussed.
2. There are some obvious formatting/English mistakes in the paper. For example:
Abstract Line11: There is an unnecessary space after the sentence "Transformer shows a good prospect in computer vision"
Introduction Line 63-66: "Section briefly reviews the work related to the Transformer model and HCCR."
Which section?
Such mistakes should be carefully identified.
3. It would be easier to understand if Figure 1 marked both the original Transformer and the Proposed S-Swin Transformer and pointed out the improvements.
In the experimental part, the author introduces the Proposed S-Swin Transformer in detail and conducts comparative experiments. Hyperparameters and database configuration are also specified in detail. However, I still feel the experiments can be made richer. In terms of the recognition success rate, the author only uses the accuracy evaluation to compare the model performance. It will be easier for the reader to understand the performance of the model if it is evaluated using the confusion matrix-based method.
Is this study validated using k-fold validation? If not, the reason needs to be briefly stated.
Overall, I think this is a very interesting study. In future work, it will be more practical to realize not only classification but also detection/segmentation.
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.