Multi-convolutional neural networks for cotton disease detection using synergistic deep learning paradigm

Afira Aslam; Syed Muhammad Usman; Muhammad Zubair; Amanullah Yasin; Muhammad Owais; Irfan Hussain

doi:10.1371/journal.pone.0324293

Peer Review History

Original SubmissionMarch 12, 2024
12 Mar 2024 Author Response https://doi.org/10.1371/journal.pone.0324293.r001
20 May 2024 Decision Letter - Jyotir Chatterjee, Editor PONE-D-24-09970Multi-Convolutional Neural Networks for Cotton Disease Detection Using Synergistic Deep Learning ParadigmPLOS ONE Dear Dr. Owais, Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. Please submit your revised manuscript by Jul 04 2024 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file. Please include the following items when submitting your revised manuscript: A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter. If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols. We look forward to receiving your revised manuscript. Kind regards, Jyotir Moy Chatterjee Academic Editor PLOS ONE Journal Requirements: When submitting your revision, we need you to address these additional requirements. 1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf 2. We suggest you thoroughly copyedit your manuscript for language usage, spelling, and grammar. If you do not know anyone who can help you do this, you may wish to consider employing a professional scientific editing service. The American Journal Experts (AJE) (https://www.aje.com/) is one such service that has extensive experience helping authors meet PLOS guidelines and can provide language editing, translation, manuscript formatting, and figure formatting to ensure your manuscript meets our submission guidelines. Please note that having the manuscript copyedited by AJE or any other editing services does not guarantee selection for peer review or acceptance for publication. Upon resubmission, please provide the following: The name of the colleague or the details of the professional service that edited your manuscript A copy of your manuscript showing your changes by either highlighting them or using track changes (uploaded as a supporting information file) A clean copy of the edited manuscript (uploaded as the new manuscript file)” 3. Please note that PLOS ONE has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, all author-generated code must be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse. 4. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match. When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section. 5. Thank you for stating the following financial disclosure: "This work was supported by KUCARS, Department of Mechanical and Nuclear Engineering, Khalifa University under Award numbers FSU-2021-019 and RC1-2018-KUCARS." Please state what role the funders took in the study. If the funders had no role, please state: ""The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."" If this statement is not correct you must amend it as needed. Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf. 6. Thank you for stating the following in the Acknowledgments Section of your manuscript: "This work was supported by KUCARS, Department of Mechanical and Nuclear Engineering, Khalifa University under Award numbers FSU-2021-019 and RC1-2018-KUCARS." We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form. Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows: "The author(s) received no specific funding for this work" Please include your amended statements within your cover letter; we will change the online submission form on your behalf. 7. We note that your Data Availability Statement is currently as follows: All relevant data are within the manuscript and its Supporting Information files. Please confirm at this time whether or not your submission contains all raw data required to replicate the results of your study. Authors must share the “minimal data set” for their submission. PLOS defines the minimal data set to consist of the data required to replicate all study findings reported in the article, as well as related metadata and methods (https://journals.plos.org/plosone/s/data-availability#loc-minimal-data-set-definition). For example, authors should submit the following data: - The values behind the means, standard deviations and other measures reported; - The values used to build graphs; - The points extracted from images for analysis. Authors do not need to submit their entire data set if only a portion of the data was used in the reported study. If your submission does not contain these data, please either upload them as Supporting Information files or deposit them to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of recommended repositories, please see https://journals.plos.org/plosone/s/recommended-repositories. If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent. If data are owned by a third party, please indicate how others may request data access. 8. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ 9. Please ensure that you refer to Figure 12 in your text as, if accepted, production will need this reference to link the reader to the figure. 10. We note you have included a table to which you do not refer in the text of your manuscript. Please ensure that you refer to Table 5 in your text; if accepted, production will need this reference to link the reader to the Table. Additional Editor Comments: Major revision [Note: HTML markup is below. Please do not edit.] Reviewers' comments: Reviewer's Responses to Questions Comments to the Author 1. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. Reviewer #1: Yes Reviewer #2: Yes ******** 2. Has the statistical analysis been performed appropriately and rigorously? Reviewer #1: Yes Reviewer #2: Yes ****** 3. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified. Reviewer #1: No Reviewer #2: Yes ****** 4. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here. Reviewer #1: Yes Reviewer #2: Yes ****** 5. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters) Reviewer #1: Authors of this manuscript addressed the problem of disease detection in cotton crop and provide the solution to detect the disease by using deep learning method. This is good piece of work but it require following corrections 1. The overall quality of the paper in general, and the implementation and presentation aspect need much improvement 2. The novelty of this research is not well explained 3. The statement “A robust system capable of addressing real-life challenges, such as diverse symptoms of diseases, occlusion, and variations in leaves” what is meaning by real life challenges is not clear. 4. Classification could be done by using SVM and random forest but LSTM is used for sequential data why all these are used collectively is not clearly explained. 5. How the stylGANs approach is addressing the problem of Data scarcity is not clear 6. Authors claimed that their method is outperforming than the existing methods but the comparison is not clear Reviewer #2: 1. The author implement state of art method like UNet or UNet++ for segmentation. 2.Fig4 should be application specific relevant to the proposed work...it looks as general figure. 3. Table 7 is not clear, the values represent what? 4. Discuss the related work in such a manner that it should highlight the method, results, advantages, and limitations. Add 8-10 more literature and discuss it properly. The final review analysis should be added at the end of this section. It will improve the readability of the paper. Add papers from 2021-2023, also Discussion section should contain the result impact, comparative study analysis and the overall analysis of the result. 5. limitations and future scope need to be included ****** 6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy. Reviewer #1: No Reviewer #2: No ******** [NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.] While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step. https://doi.org/10.1371/journal.pone.0324293.r002
Revision 1
28 Jun 2024 Author Response < Editor Comments > Dear Editor, We appreciate your review of our submitted manuscript titled “Multi-Convolutional Neural Networks for Cotton Disease Detection Using Ensemble Learning” in the PLOS One journal. We have thoroughly revised our paper according to the “Journal Requirements” and the comments provided by the “Reviewers”. All the revisions are highlighted in red and blue text in the revised manuscript. We have also included a rebuttal to address the changes made in response to each reviewer's comments. Best Regards, Muhammad Owais (Ph.D.) Corresponding author muhammad.owais@ku.ac.ae Contanct#: +971-58-135-8344 < Journal Requirements > < Comment 1 > Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf < Response > We confirm that our manuscript meets PLOS ONE’s style requirements, including those for file naming. < Comment 2 > We suggest you thoroughly copyedit your manuscript for language usage, spelling, and grammar. If you do not know anyone who can help you do this, you may wish to consider employing a professional scientific editing service. The American Journal Experts (AJE) (https://www.aje.com/) is one such service that has extensive experience helping authors meet PLOS guidelines and can provide language editing, translation, manuscript formatting, and figure formatting to ensure your manuscript meets our submission guidelines. Please note that having the manuscript copyedited by AJE or any other editing services does not guarantee selection for peer review or acceptance for publication. Upon resubmission, please provide the following: The name of the colleague or the details of the professional service that edited your manuscript A copy of your manuscript showing your changes by either highlighting them or using track changes (uploaded as a supporting information file) A clean copy of the edited manuscript (uploaded as the new manuscript file)” < Response > Prof. Dr. Shehzad Khalid proofread the manuscript. Both clean copy and track changes are being submitted. < Comment 3 > Please note that PLOS ONE has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, all author-generated code must be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse. < Response > The authors agree to share the code to reproduce the results of this research. < Comment 4 > We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match. When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section. < Response > Corrected: “This work was supported by KUCARS, Department of Mechanical and Nuclear Engineering, Khalifa University under Award numbers FSU-2021-019 and RC1-2018-KUCARS.” < Comment 5 > Thank you for stating the following financial disclosure: "This work was supported by KUCARS, Department of Mechanical and Nuclear Engineering, Khalifa University under Award numbers FSU-2021-019 and RC1-2018-KUCARS." Please state what role the funders took in the study. If the funders had no role, please state: ""The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."" If this statement is not correct you must amend it as needed. Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf. < Response > We amended the role of funder statement and also included in cover letter in yellow highlighted text. “Role of Funders: The funders oversaw this research and helped in preparing the manuscript.” < Comment 6 > Thank you for stating the following in the Acknowledgments Section of your manuscript: "This work was supported by KUCARS, Department of Mechanical and Nuclear Engineering, Khalifa University under Award numbers FSU-2021-019 and RC1-2018-KUCARS." We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form. Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows: "The author(s) received no specific funding for this work" Please include your amended statements within your cover letter; we will change the online submission form on your behalf. < Response > We have renamed the “Funding and Acknowledgement” section to “Funding Statement” in the revised manuscript and updated the funding statement as follows: Previous statement: The author(s) received no specific funding for this work. Revised statement: This work was supported by KUCARS, Department of Mechanical and Nuclear Engineering, Khalifa University under Award numbers FSU-2021-019 and RC1-2018-KUCARS. < Comment 7 > We note that your Data Availability Statement is currently as follows: All relevant data are within the manuscript and its Supporting Information files. Please confirm at this time whether or not your submission contains all raw data required to replicate the results of your study. Authors must share the “minimal data set” for their submission. PLOS defines the minimal data set to consist of the data required to replicate all study findings reported in the article, as well as related metadata and methods (https://journals.plos.org/plosone/s/data-availability#loc-minimal-data-set-definition). For example, authors should submit the following data: - The values behind the means, standard deviations and other measures reported; - The values used to build graphs; - The points extracted from images for analysis. Authors do not need to submit their entire data set if only a portion of the data was used in the reported study. If your submission does not contain these data, please either upload them as Supporting Information files or deposit them to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of recommended repositories, please see https://journals.plos.org/plosone/s/recommended-repositories. If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent. If data are owned by a third party, please indicate how others may request data access. < Response > In this study, we used publicly available open datasets. These publicly available datasets links are provided below: Data source links: https://www.kaggle.com/datasets/seroshkarim/cotton-leaf-disease-dataset https://data.mendeley.com/datasets/6hm6pc3y43/2 https://universe.roboflow.com/national-college-of-ireland/cotton-plant-disease-prediction-igthk/dataset/3 https://universe.roboflow.com/roboflow-100/cotton-plant-disease/dataset/2 There are no such graphs that used mean, standard deviation, etc., values to plot. However, to reproduce the research work authors have included the details on the data augmentation (Table 3), model details (Table 4, 5, and 8). < Comment 8 > PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ < Response > Thanks for your update. The corresponding author has linked his ORCID with the editorial manager. https://orcid.org/0000-0001-7679-081X < Comment 9 > Please ensure that you refer to Figure 12 in your text as, if accepted, production will need this reference to link the reader to the figure. < Response > Figure 12 has been referenced in the text and updated in the revised manuscript. Please see 2nd paragraph of “Results” section. < Comment 10 > We note you have included a table to which you do not refer in the text of your manuscript. Please ensure that you refer to Table 5 in your text; if accepted, production will need this reference to link the reader to the Table. < Response > Authors would like to thank the editor and have referenced the Table 5 in the text. Please see the “Proposed Methodology” section, the paragraph starting with the following statement: Table 5 provides detail of parameters of MobileNet V2. < Reviewer 1 > < General Comment > Authors of this manuscript addressed the problem of disease detection in cotton crop and provide the solution to detect the disease by using deep learning method. This is good piece of work but it require following corrections. < Response > Thanks for your positive response and valuable comments. We have thoroughly revised our paper according to the comments. All the revisions are highlighted in red and blue text in the revised manuscript. < Comment 1 > The overall quality of the paper in general, and the implementation and presentation aspect need much improvement. < Response > We have thoroughly revised our paper by addressing the implementation and result presentation to improve the overall quality of the paper. < Comment 2 > The novelty of this research is not well explained < Response > We apologize for the confusion. We have thoroughly revised the contribution and discussion sections of our paper to highlight the novelty of this research as follows: Novelty of the study: We proposed a comprehensive system designed to effectively manage and overcome these challenges, ensuring reliable and accurate performance in diverse agricultural environments. Contributions of this research to the body of knowledge include: • Developed a deep ensemble method to accurately classify various cotton diseases, addressing challenges of class imbalance and symptom variation through a new data augmentation scheme and feature fusion approach. • Proposed a synthetic data augmentation scheme using a customized StyleGAN along with conventional methods (scaling, rotating, transforming, shearing, zooming) to tackle class imbalance. • Introduced a dual-stage fusion scheme by combining features extracted from MobileNet and VGG16 to create a comprehensive feature vector, followed by a StackNet-based ensemble classifier integrating Long and Short Term Memory (LSTM), Support Vector Machine (SVM), and Random Forest for enhanced prediction accuracy. • Achieved an average accuracy of 97% on a publicly available dataset, surpassing state-of-the-art techniques in identifying six diseases and a healthy class. These modifications are included at the end of the “Introduction” section. Key Findings and Potential Significance: The proposed ensemble method, trained with categorical cross-entropy loss, RMSProp optimizer, and softmax activation, achieves a 97\% accuracy. This research introduces several novel aspects of precision agriculture, particularly in detecting and managing crop diseases and developing a robust method that deals with real-world challenges such as varying disease symptoms, occlusion from overlapping leaves, and diverse leaf characteristics in shape, size, and texture. A novel data augmentation technique has been developed to address the issue of data scarcity and class imbalance in agricultural datasets. Incorporating multi-convolutional neural networks combined with ensemble classification techniques is a novel approach to improve the accuracy of disease detection. This method combines the strengths of convolutional neural networks for feature extraction and classification. As a result, it accurately identifies six specific diseases and a healthy class in cotton crops. The above paragraph highlighting the novelty of this study is also added as the last paragraph in the “Discussion” section. < Comment 3 > “A robust system capable of addressing real-life challenges, such as diverse symptoms of diseases, occlusion, and variations in leaves” what is meaning by real life challenges is not clear < Response > We apologize for the vague explanation. We revised these lines as follows: A robust deep learning system is essential for addressing the practical challenges encountered in real-world agricultural settings. These challenges include diverse symptoms of diseases, such as varying sizes and colors of lesions, which can complicate accurate diagnosis. Additionally, occlusion, where leaves are partially covered by other leaves or objects, poses a significant obstacle to clear image capture and analysis. Variations in leaf appearance, including differences in shapes, sizes, and textures, further add to the complexity. To tackle these issues, we proposed a comprehensive system designed to effectively manage and overcome these challenges, ensuring reliable and accurate performance in diverse agricultural environments. These lines are updated at the end of the “Introduction” section (before the 1st bullet point of the contribution part). < Comment 4 > Classification could be done by using SVM and random forest but LSTM is used for sequential data why all these are used collectively is not clearly explained. < Response > Thanks for your valuable comment. We added further details to the proposed methodology to explain the significance of LSTM, which is used along with SVM and RF as followas: The use of LSTM, despite its typical application for sequential data, is beneficial as it captures temporal dependencies or patterns within the feature vectors that traditional classifiers like SVM and Random Forest might overlook. By integrating these diverse methods, the strength of LSTM in handling sequential patterns is leveraged alongside the robustness of Random Forest in managing non-linear relationships and the effectiveness of SVM in high-dimensional spaces, thereby enhancing overall model performance. Please see the 2nd paragraph of the section “Proposed Methodology”. < Comment 5 > How the stylGANs approach is addressing the problem of Data scarcity is not clear. < Response > Explanation is added to clarify how styleGANs help with data scarcity as follows: StyleGANs address the problem of data scarcity by generating high-quality synthetic images that are indistinguishable from real images. This increases the dataset size and ensures diversity and variety in the augmented data, thus enhancing the model's ability to generalize better. Please see the 3rd paragraph of the section “Proposed Methodology”. < Comment 6 > Authors claimed that their method is outperforming than the existing methods but the comparison is not clear < Response > We apologize for the vague explanation. We have thoroughly revised the comparison table and highlighted the performance gains of the proposed method compared to existing methods as follows: Comparison Table 8 presents a comparative a Attachments Attachment Submitted filename: 2024-06-22-Response to Editor and Reviewers.docx https://doi.org/10.1371/journal.pone.0324293.r003
8 Aug 2024 Decision Letter - Jyotir Chatterjee, Editor PONE-D-24-09970R1Multi-Convolutional Neural Networks for Cotton Disease Detection Using Synergistic Deep Learning ParadigmPLOS ONE Dear Dr. Owais, Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. Please submit your revised manuscript by Sep 22 2024 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file. Please include the following items when submitting your revised manuscript: A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter. If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols. We look forward to receiving your revised manuscript. Kind regards, Jyotir Moy Chatterjee Academic Editor PLOS ONE Additional Editor Comments: Major Revision [Note: HTML markup is below. Please do not edit.] Reviewers' comments: Reviewer's Responses to Questions Comments to the Author 1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation. Reviewer #1: All comments have been addressed Reviewer #3: All comments have been addressed ******** 2. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. Reviewer #1: Yes Reviewer #3: Partly ****** 3. Has the statistical analysis been performed appropriately and rigorously? Reviewer #1: Yes Reviewer #3: Yes ****** 4. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified. Reviewer #1: Yes Reviewer #3: No ****** 5. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here. Reviewer #1: Yes Reviewer #3: Yes ****** 6. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters) Reviewer #1: The authors edited the manuscript according to the comments. Therefore accepted for the publication. Reviewer #3: The authors have not been able to show the new contribution. This paper does not add anything new to knowledge as the authors dont state the major contributions of the work. There is no comparative analysis with the state-of-the-art methods. ****** 7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy. Reviewer #1: No Reviewer #3: No ******** [NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.] While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step. https://doi.org/10.1371/journal.pone.0324293.r004
Revision 2
17 Aug 2024 Author Response < Reviewer 1 > < Comment > The authors edited the manuscript according to the comments. Therefore, accepted for the publication. < Response > We would like to thank the reviewer and appreciate your reviewing process of our manuscript. < Reviewer 3 > < Comment 1 > The authors have not been able to show the new contribution. This paper does not add anything new to knowledge as the authors dont state the major contributions of the work. < Response > Authors would like to thank the reviewer for the valuable comment. We have highlighted the contributions and incorporated them in the Introduction section, lines 48-57 on page no. 3. Action taken in the revised manuscript • Developed a novel method for data augmentation by combining conventional methods and custom Style GANs to reduce class imbalance problems. • Introduced feature-level fusion scheme by combining features extracted from MobileNet and VGG16 to create a comprehensive feature vector • Proposed an ensemble classifier using StackNet by integrating Long Short Term Memory Units (LSTM), Support Vector Machine (SVM), and Random Forest to classify cotton crop diseases accurately. • Achieved an average accuracy of 97\% on publicly available datasets, surpassing state-of-the-art techniques in classifying seven classes, including six diseases and a healthy class. < Comment 2 > There is no comparative analysis with the state-of-the-art methods. < Response > The authors would like to thank the reviewer for the valuable suggestion. We have added a comparison of the proposed method with existing state-of-the-art methods. Table 8 has been updated in the results section on page no. 17 and described in the text. Action taken in the revised manuscript Table 8 presents a comparative analysis of the proposed model versus existing methods for cotton disease detection, focusing on their parameters (#Par in millions) and accuracy. The proposed ensemble model achieves the highest accuracy of 0.97, indicating a 6% performance gain over the next best models. The second-best models [50, 54], with accuracies of 0.91 and 0.90, include MobileNetV2 with a learning scheduler (0.91, 3.5M), MobileNetV2 with more dense layers (0.90, 3.5M), and EfficientNet (0.90, 5.3M). The third-best models [50, 51], with an accuracy of 0.88, include MobileNetV2 (3.5M), MobileNetV2 unfreezing initial layers (3.5M) and Xception (22.9M). Despite the proposed method’s high parameter count (143.4M), its 6% performance gain justifies the complexity, significantly outperforming models like MobileNetV2 (3.5M) and DenseNet121 (8.0M), which achieve relatively high accuracy. This highlights the effectiveness of our ensemble model in achieving superior accuracy, balancing parameter size and performance gain, and demonstrating the model’s efficacy in cotton disease detection applications. Attachments Attachment Submitted filename: 2024-08-17-Response to Editor and Reviewers.pdf https://doi.org/10.1371/journal.pone.0324293.r005
27 Dec 2024 Decision Letter - Jyotir Chatterjee, Editor PONE-D-24-09970R2Multi-Convolutional Neural Networks for Cotton Disease Detection Using Synergistic Deep Learning ParadigmPLOS ONE Dear Dr. Owais, Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. ============================== ACADEMIC EDITOR: Minor Revision============================== Please submit your revised manuscript by Feb 10 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file. Please include the following items when submitting your revised manuscript: A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter. If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols. We look forward to receiving your revised manuscript. Kind regards, Jyotir Moy Chatterjee Academic Editor PLOS ONE Journal Requirements: Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice. [Note: HTML markup is below. Please do not edit.] Reviewers' comments: Reviewer's Responses to Questions Comments to the Author 1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation. Reviewer #4: All comments have been addressed Reviewer #5: All comments have been addressed ******** 2. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. Reviewer #4: Yes Reviewer #5: Partly ****** 3. Has the statistical analysis been performed appropriately and rigorously? Reviewer #4: Yes Reviewer #5: Yes ****** 4. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified. Reviewer #4: Yes Reviewer #5: Yes ****** 5. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here. Reviewer #4: Yes Reviewer #5: Yes ****** 6. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters) Reviewer #4: (No Response) Reviewer #5: Comments: Table 8 - Major Revision Needed: Table 8 presents important findings that compare various methods and their associated accuracies. However, there is significant ambiguity in the dataset used and some discrepancies with the reported accuracies. Dataset Consistency: It is crucial to clarify whether all the studies referenced in Table 8 are based on the same dataset. The manuscript does not specify whether these methods are evaluated using identical datasets or if different datasets were used across the studies. This is essential for interpreting the accuracy comparisons accurately. I recommend adding a note to Table 8 or in the figure legend specifying whether the methods are evaluated on the same dataset, and if not, providing details about the datasets used for each study. Inconsistencies in Reported Accuracy: The manuscript mentions that the accuracy of the method from Reference 53 is 86%, but the published version of Reference 53 states that the accuracy of this method is 96.46%. This discrepancy needs to be addressed. Please confirm the accuracy values, and if there was an error in reporting, correct it in both the manuscript and Table 8. Similarly, the manuscript states that the accuracy for the method from Reference 50 is 0.88, but the published copy of Reference 50 does not report this value as accuracy. Instead, 0.88 refers to the Area Overlap Measure (AOM), which is a different metric. The manuscript should correct this mistake and clarify that 0.88 corresponds to the AOM, not accuracy. ****** 7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy. Reviewer #4: No Reviewer #5: No ******** [NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.] While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step. https://doi.org/10.1371/journal.pone.0324293.r006
Revision 3
5 Feb 2025 Author Response REVIEWER – 5 Comments: Table 8 - Major Revision Needed: Table 8 presents important findings that compare various methods and their associated accuracies. However, there is significant ambiguity in the dataset used and some discrepancies with the reported accuracies. Dataset Consistency: It is crucial to clarify whether all the studies referenced in Table 8 are based on the same dataset. The manuscript does not specify whether these methods are evaluated using identical datasets or if different datasets were used across the studies. This is essential for interpreting the accuracy comparisons accurately. I recommend adding a note to Table 8 or in the figure legend specifying whether the methods are evaluated on the same dataset, and if not, providing details about the datasets used for each study. Inconsistencies in Reported Accuracy: The manuscript mentions that the accuracy of the method from Reference 53 is 86%, but the published version of Reference 53 states that the accuracy of this method is 96.46%. This discrepancy needs to be addressed. Please confirm the accuracy values, and if there was an error in reporting, correct it in both the manuscript and Table 8. Similarly, the manuscript states that the accuracy for the method from Reference 50 is 0.88, but the published copy of Reference 50 does not report this value as accuracy. Instead, 0.88 refers to the Area Overlap Measure (AOM), which is a different metric. The manuscript should correct this mistake and clarify that 0.88 corresponds to the AOM, not accuracy. Response: The authors would like to thank the reviewer for the valuable comment. We appreciate the reviewer's valuable feedback regarding the dataset consistency and reported accuracy values in Table 8. Based on the suggestions provided, we have made the necessary clarifications and corrections to enhance the accuracy and transparency of the reported findings. Dataset Consistency Clarification: We acknowledge the reviewer's concern regarding dataset consistency across the referenced studies. As suggested, we have explicitly added a new column to Table 8, specifying the datasets used in each study along with their respective references. This addition ensures clarity regarding the sources of data and allows for a more accurate interpretation of the accuracy comparisons. Furthermore, we have updated the table caption to explicitly state that the studies utilize different datasets for either binary or multiclass classification. These modifications directly address the reviewer's request for transparency in dataset usage. Corrections to Reported Accuracy Values: We have carefully reviewed and rectified any discrepancies in the reported accuracy values. Specifically: • The accuracy value for Reference 53 was previously incorrect and has now been corrected to match the published value (0.96). • The value for Reference 50 is corrected with (0.91) for MobileNet V2 and (0.98) with added dense layers. • In addition to these corrections, we have thoroughly counterchecked the accuracy and parameter values for all other studies in the table to ensure accuracy and consistency. Any discrepancies identified during this process have been corrected accordingly. Action: Please find in the attached response file https://doi.org/10.1371/journal.pone.0324293.r007
10 Mar 2025 Decision Letter - Jyotir Chatterjee, Editor PONE-D-24-09970R3Multi-Convolutional Neural Networks for Cotton Disease Detection Using Synergistic Deep Learning ParadigmPLOS ONE Dear Dr. Owais, Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. ============================== ACADEMIC EDITOR: ============================== Please submit your revised manuscript by Apr 24 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file. Please include the following items when submitting your revised manuscript: A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter. If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols. We look forward to receiving your revised manuscript. Kind regards, Jyotir Moy Chatterjee Academic Editor PLOS ONE Additional Editor Comments: 1. The study lacks clarity on whether all referenced methods in Table 8 are evaluated using the same dataset. This makes it difficult to accurately compare accuracy metrics across different approaches. 2. Discrepancies exist in reported accuracy values for references. For example, the manuscript previously stated that Reference 53 had an accuracy of 86%, but the published value was 96.46%. Similarly, Reference 50's accuracy was misrepresented. 3. The model is trained in a controlled setting, which may not accurately reflect real-world variability, including lighting conditions, environmental noise, and different imaging devices. This could lead to reduced generalizability when deployed in practical scenarios. 4. Although the study attempts to address class imbalance through synthetic data generation (StyleGAN and augmentation), reliance on synthetic data might introduce biases that do not accurately represent real-world variations in disease symptoms. 5. The reported 97% accuracy suggests a potential overfitting issue, especially given that the dataset is relatively small and augmented artificially. More robust validation techniques, such as external dataset testing, would strengthen the claims. 6. The study does not provide sufficient interpretability techniques (e.g., Grad-CAM, SHAP) to understand why the model makes certain predictions. Explainable AI methods could enhance trust in the system for agricultural experts and farmers. [Note: HTML markup is below. Please do not edit.] Reviewers' comments: Reviewer's Responses to Questions Comments to the Author 1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation. Reviewer #6: (No Response) ******** 2. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. Reviewer #6: Yes ****** 3. Has the statistical analysis been performed appropriately and rigorously? Reviewer #6: Yes ****** 4. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified. Reviewer #6: Yes ****** 5. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here. Reviewer #6: Yes ****** 6. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters) Reviewer #6: I found this to be a solid paper on cotton disease detection. The authors' approach of combining features from MobileNet and VGG16, then using an ensemble classifier (StackNet with LSTM, SVM, and Random Forest) is creative and effective. What works well: Their method addresses real challenges in field conditions, particularly variation in disease symptoms The StyleGAN for synthetic data generation is a clever solution to class imbalance Using the Segment Anything Model for disease area quantification adds practical value A few suggestions that might strengthen the paper: I'd like to see more about inference time and computational requirements, since that matters for field deployment While they do compare parameter counts (143.4M is quite large), some discussion about deployment on resource-limited devices would be helpful More testing on challenging field conditions (varied lighting, partial leaf occlusion) would better demonstrate robustness The corrections they made to Table 8 in response to Reviewer 5 have addressed those concerns appropriately. Overall, this represents a meaningful contribution that advances cotton disease detection methods. ****** 7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy. Reviewer #6: No ******** [NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.] While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step. https://doi.org/10.1371/journal.pone.0324293.r008
Revision 4
20 Apr 2025 Author Response Dear Editor, We appreciate your reviewing process of our submitted manuscript titled “Multi-Convolutional Neural Networks for Cotton Disease Detection Using Ensemble Learning” in the PLOS One journal. We have revised our paper according to the comments provided by the editor and reviewers. All added text is in Red, with removed text as (strike through) in the “Revised Manuscript with Track Changes”. Moreover, a “Response to Reviewer(s)” file is also provided to address the changes made in the revised paper against individual comments by the reviewers. In addition, a clean file “Manuscript” is also uploaded in separate. Best regards Editor Comments Comment 1: The study lacks clarity on whether all referenced methods in Table 8 are evaluated using the same dataset. This makes it difficult to accurately compare accuracy metrics across different approaches. Response: We appreciate the reviewer’s observation. In Table 8 of the original manuscript, we presented the results of prior studies primarily based on the "PlantVillage" dataset. However, our proposed model was evaluated on a combined dataset comprising both the "PlantVillage" and "Roboflow" datasets. We acknowledge that this discrepancy may have made cross-study comparisons unclear. To address this, we have revised Table 8 to include recent studies that report performance on the "Roboflow" dataset as well. This ensures a more accurate and meaningful comparison. While two existing studies report slightly higher accuracy on the "PlantVillage" dataset alone, it is important to highlight that our model demonstrates robust performance across both datasets when combined, suggesting better generalization capability. Moreover, we have removed the result of one study which is shows results on a completely different dataset. Our model’s strong performance on the combined datasets “PlantVillage” and “Roboflow” datasets is considered more challenging, further emphasizes its effectiveness and practical applicability. Action: Please see the section “Comparison with State-of-the-Art methods” and Table 8 on page numbers 16-17 in the revised track changes manuscript. Comment 2: Discrepancies exist in reported accuracy values for references. For example, the manuscript previously stated that Reference 53 had an accuracy of 86%, but the published value was 96.46%. Similarly, Reference 50's accuracy was misrepresented. Response: We appreciate the reviewer for pointing out these discrepancies. Upon review, we found that Reference 53 was based on a different dataset that was not used in our model evaluation. To maintain a fair and consistent comparison, we have removed this reference from Table 8, as now reflected in the revised manuscript. Additionally, the accuracy value for Reference 50 has been corrected accordingly. We have also thoroughly verified all other referenced accuracy values to ensure consistency and accuracy across the manuscript. Action: Please see Table 8 on page numbers 17 in the revised track changes manuscript. Comment 3: The model is trained in a controlled setting, which may not accurately reflect real-world variability, including lighting conditions, environmental noise, and different imaging devices. This could lead to reduced generalizability when deployed in practical scenarios. Response: We acknowledge the reviewer’s concern and have addressed this by evaluating our model on a combined dataset comprising PlantVillage and Roboflow, which differ significantly in terms of image quality, lighting conditions, background complexity, and noise. This combination introduces higher data variability and better reflects real-world scenarios. Unlike single-dataset studies, our approach provides a more rigorous test of generalization. Notably, our model performed strongly on the more challenging Roboflow dataset, demonstrating its robustness and applicability in diverse practical settings. Action: Please see the added text on page number 17 in the revised track changes manuscript. Comment 4: Although the study attempts to address class imbalance through synthetic data generation (StyleGAN and augmentation), reliance on synthetic data might introduce biases that do not accurately represent real-world variations in disease symptoms. Response: We appreciate the reviewer’s valuable observation. While synthetic data generation using StyleGAN and augmentation was used to address class imbalance, we recognize the potential risk of introducing biases that may not fully capture the complexity of real-world disease manifestations. To mitigate this, we ensured that the synthetic samples were only used during the training phase and not in validation or testing, which were conducted exclusively on real images from the combined PlantVillage and Roboflow datasets. Action: See 2nd paragraph of the “Proposed Methodology” on page number 8 in the revised track changes manuscript. Comment 5: The reported 97% accuracy suggests a potential overfitting issue, especially given that the dataset is relatively small and augmented artificially. More robust validation techniques, such as external dataset testing, would strengthen the claims. Response: We thank the reviewer for raising this important concern. To address the potential overfitting issue, we conducted an external validation using a completely unseen, publicly available cotton leaf disease dataset [74]. Unlike prior studies that used parts of this dataset for both training and testing, our model was evaluated solely on the test set without any prior exposure. Despite this, our model achieved a competitive accuracy of 96.4%, confirming its generalization ability. These results, combined with strong performance on diverse datasets like PlantVillage and Roboflow, demonstrate the robustness and real-world applicability of our model. We have included this external validation and discussion in the revised manuscript (see Section "External Validation Testing" and Table 9). Moreover, the sections “Discussion”, “Key Findings and Potential Significance”, “Limitations and Future Directions”, and “Conclusion” are also updated accordingly. Action: See section “External Dataset Testing” on page number 18, Table 9 on page 19, and “Discussion” section on page 19 and 21, “Key Findings and Potential Significance” section on page 21, “Limitations and Future Directions” on page 21-22, and “Conclusion” section on page 22 in the revised track changes manuscript. Discussion Key Findings and Potential Significance Limitations and Future Directions Conclusion Comment 6: The study does not provide sufficient interpretability techniques (e.g., Grad-CAM, SHAP) to understand why the model makes certain predictions. Explainable AI methods could enhance trust in the system for agricultural experts and farmers. Response: We appreciate the reviewer’s valuable suggestion. In the revised manuscript, we have incorporated the Explainable AI technique Grad-CAM to provide visual explanations for the model's predictions. A new section titled “Visual Explanation of Model Decisions” has been added, along with Figure 9, which presents Grad-CAM visualizations for VGG-16, MobileNet V2, StackNet, and our proposed ensemble model. These visualizations highlight the discriminative regions used by each model, thereby improving transparency and helping end-users understand and trust the model’s decision-making process. Action: See section “Visual Explanation of Model Decisions” on page 19 and Figure 9 on page 20 in the revised track changes manuscript. Attachments Attachment Submitted filename: Response to Reviewers.pdf https://doi.org/10.1371/journal.pone.0324293.r009
23 Apr 2025 Decision Letter - Jyotir Chatterjee, Editor Multi-Convolutional Neural Networks for Cotton Disease Detection Using Synergistic Deep Learning Paradigm PONE-D-24-09970R4 Dear Dr. Owais, We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements. Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication. An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. If you have any questions relating to publication charges, please contact our Author Billing department directly at authorbilling@plos.org. If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org. Kind regards, Jyotir Moy Chatterjee Academic Editor PLOS ONE Additional Editor Comments (optional): Reviewers' comments: https://doi.org/10.1371/journal.pone.0324293.r010
Formally Accepted
Acceptance Letter - Jyotir Chatterjee, Editor PONE-D-24-09970R4 PLOS ONE Dear Dr. Owais, I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now being handed over to our production team. At this stage, our production department will prepare your paper for publication. This includes ensuring the following: * All references, tables, and figures are properly cited * All relevant supporting information is included in the manuscript submission, * There are no issues that prevent the paper from being properly typeset You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps. Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org. If we can help with anything else, please email us at customercare@plos.org. Thank you for submitting your work to PLOS ONE and supporting open access. Kind regards, PLOS ONE Editorial Office Staff on behalf of Mr. Jyotir Moy Chatterjee Academic Editor PLOS ONE https://doi.org/10.1371/journal.pone.0324293.r011

Open letter on the publication of peer review reports

PLOS recognizes the benefits of transparency in the peer review process. Therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. Reviewers remain anonymous, unless they choose to reveal their names.

We encourage other journals to join us in this initiative. We hope that our action inspires the community, including researchers, research funders, and research institutions, to recognize the benefits of published peer review reports for all parts of the research system.

Learn more at ASAPbio .