Application of Target Detection Based on Deep Learning in Intelligent Mineral Identification
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsLine 184: reference more papers that used this method even in other fields. What are the main applications of the method you used?
As a paper in the application of AI it is better to just mention the algorithm and explain it in summary this is not a new AI method better to keep the background of the algorithm shorter and focus more on the geology aspects of it with a deeper explanation. People using AI know many of the backgrounds and you can reference the papers you used for their use. Better to focus on the novelty of your work more.
The purpose of the paper is unclear. Is this going to be used for museums? Are people spending a lot of time and cost detecting the minerals? who would benefit from this? People need to crop mineral images and separate them from the background in a museum? Please explain in a more clear way.
The test set is not representative of the performance of the model. You are using 1300 images for train and validation and only testing 15 images? Based on your data size, you should go with train/validation/ test of either 60/20/20 or 70/15/15 percent of each class. You are using only 1%!! Although you had different numbers of samples for different classes you only considered 15 images for the test set for all of them which is not a good validation of your model.
Again in section 4, you explained very basic background either remove them or summarize. People using AI already know these topics you should explain your application more not the basics of Machine learning
Provide a table with the results of your test set.
More detailed explanation in the conclusion section.
Comments on the Quality of English LanguageCan be improved and the sentences can be stated more clearly.
Author Response
COMMENT 1: Line 184: reference more papers that used this method even in other fields. What are the main applications of the method you used?
RESPONE 1: Thank you for your insightful feedback. In response to your suggestion, we have revised the manuscript to include additional references that illustrate the broad applications of the YOLO algorithm across various fields. These references highlight the versatility and robustness of the YOLO algorithm, which supports its use as a core tool in our study on intelligent mineral recognition. We have also expanded the discussion in the manuscript to provide a detailed overview of the main applications of the YOLO algorithm. These additions aim to provide a clearer understanding of why this method was chosen for our research and how it has been successfully applied in other domains. Changes Made:
1, We have added several references to studies from different disciplines that have successfully utilized the YOLO algorithm, as detailed below:
Biology: Abdullah et al. (2022) [45] used YOLO for real-time detection of fish species in underwater videos.
Traffic Monitoring: Zuraim et al. (2021) [46] applied YOLO in traffic surveillance systems for vehicle detection and tracking.
Agriculture: Vilar-Andreu et al. (2024) [47] developed a pest detection system for crops using YOLO.
Medical Imaging: Prinzi et al. (2024) [48] employed YOLO to detect tumors in mammograms.
Industrial Applications: Reddy et al. (2024) [49] used YOLO in semiconductor manufacturing for defect detection.
[These revisions have been included in the fifth major paragraph of chapter 1 of the manuscript]
2, The main applications of the method we used is YOLO V8-x, which had mentioned in Section 4.1. We hope these changes meet the reviewer's expectations and enhance the clarity and comprehensiveness of our work. Thank you for your constructive comments and suggestions.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 2: As a paper in the application of AI it is better to just mention the algorithm and explain it in summary this is not a new AI method better to keep the background of the algorithm shorter and focus more on the geology aspects of it with a deeper explanation. People using AI know many of the backgrounds and you can reference the papers you used for their use. Better to focus on the novelty of your work more.
RESPONE 2:
Thank you for your insightful feedback on our manuscript. We appreciate your suggestions and have revised the paper accordingly.
In response to your comments, we have condensed the background information on the AI algorithm to focus more on its application within the geological context of our study. We have reduced the detailed explanation of the algorithm and instead referenced the relevant literature where readers can find more comprehensive background information. [These revisions have been included in the paragraph of chapter 2 of the manuscript]
We have also expanded the section on the geological aspects of our work to provide a deeper explanation of its significance and novelty. This includes a more detailed discussion of how our application of the algorithm contributes to advancements in geological research. [These revisions have been included in the paragraph of chapter 6 of the manuscript]
We hope these revisions address your concerns and enhance the clarity and impact of our manuscript. Thank you again for your valuable feedback.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 3: The purpose of the paper is unclear. Is this going to be used for museums? Are people spending a lot of time and cost detecting the minerals? who would benefit from this? People need to crop mineral images and separate them from the background in a museum? Please explain in a more clear way.
RESPONE 3: Thank you for your valuable feedback and for raising important questions regarding the purpose and application of our study. We appreciate the opportunity to clarify these aspects. The primary aim of our paper is to develop and demonstrate an AI-based method for mineral detection and classification. This method is designed to enhance the efficiency and accuracy of mineral analysis, which can be beneficial in various contexts, including both research and practical applications. To address your specific queries:
1, Purpose and Application: Our method is not specifically tailored for museum use but rather for broader applications in mineralogy and geological research. However, it could indeed be useful for museums in the context of cataloging and analyzing mineral specimens. The core aim is to provide a more efficient and accurate tool for mineral detection, which can be applied in various settings, including research labs, educational institutions, and potentially museums.
2, Cost and Time Efficiency: Currently, detecting and classifying minerals can be time-consuming and costly, particularly when done manually. Our AI-based approach aims to reduce both time and costs associated with these tasks by automating the detection process and improving accuracy. It does not require people to cut images before identification, but can be directly connected to the computer's camera for real-time target detection.
3, Target Users: The primary beneficiaries of our method include geologists, researchers, and educators who need to analyze large collections of mineral samples. Additionally, it could be of interest to institutions involved in mineral preservation and curation.
4, Image Processing Needs: While our method can be applied to mineral images from various sources, it is particularly valuable in situations where mineral images need to be processed, and their features separated from the background. This capability is relevant not only for museum contexts but also for any scenario where high-quality image analysis of mineral specimens is required.
We hope this addresses your concerns and provides a clearer understanding of the purpose and impact of our study. Thank you again for your constructive feedback.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 4: The test set is not representative of the performance of the model. You are using 1300 images for train and validation and only testing 15 images? Based on your data size, you should go with train/validation/ test of either 60/20/20 or 70/15/15 percent of each class. You are using only 1%!! Although you had different numbers of samples for different classes you only considered 15 images for the test set for all of them which is not a good validation of your model.
RESPONE 4: Thank you for your constructive feedback regarding the test set size in our study. We appreciate your insights and have taken them into account to improve our model evaluation. In response to your concerns, we have made significant revisions to the test set as follows:
1, Expanded Test Set: We have increased the number of test images for each mineral category. Originally, each class had only 15 test images, which was insufficient for a robust evaluation. We have now supplemented each class with additional images, bringing the total number of images per class to 50.
2, Augmentation for Increased Diversity: To further enhance the representativeness of the test set, we applied various data augmentation techniques, including rotation and mirroring. This augmentation has expanded the number of images per class from 50 to 200, resulting in a total of 1400 images in the test set.
3, Revised Evaluation: With this expanded and augmented test set, we have re-evaluated the performance of our model. The revised results, including detailed performance metrics, are now included in the manuscript. This updated evaluation provides a more comprehensive assessment of the model’s accuracy and robustness across different mineral categories.
[These revisions have been included in the paragraph of chapter 3.1 of the manuscript]
We believe these improvements address the issues raised and provide a more thorough and representative validation of our model. Thank you for your valuable feedback, which has helped us enhance the quality and reliability of our study.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 5: Again in section 4, you explained very basic background either remove them or summarize. People using AI already know these topics you should explain your application more not the basics of Machine learning.
RESPONE 5: Thank you for your insightful feedback on Section 4 of our manuscript. We appreciate your suggestions and have revised the section accordingly. We understand that the background information provided in Section 4 may be too basic for readers who are already familiar with machine learning concepts. In response to your comments, we have made the following adjustments:
1, Condensed Background Information: We have removed the basic explanations of machine learning concepts that may be redundant for readers with a background in AI. The section now includes only a brief summary of the fundamental concepts necessary to understand the application of our method.
2, Enhanced Focus on Application: We have significantly expanded the discussion on the specific application of our AI method within the context of our study. This includes a detailed explanation of how our approach is tailored to address the unique challenges of mineral detection and classification, and how it improves upon existing methods.
3, Emphasized Novelty and Contribution: The revised section highlights the novelty of our work, including any new techniques or insights that our approach offers. We have included more details on the practical implications and benefits of our method, making it clear how our application advances the field.
[These revisions have been included in the paragraph of chapter 4.1 of the manuscript]
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 6: Provide a table with the results of your test set.
RESPONE 6: Thank you for your feedback regarding the inclusion of test set results. We appreciate your suggestion and have addressed it by providing a detailed summary of the test set results. In response to your request, we have included the following results in the manuscript:
The confidence levels for each mineral class in the test set are as follows: Black Mica: 85.25%, Quartz: 83.12%, Chalcopyrite: 82.30%, Siliceous Malachite: 80.08%, Malachite: 79.82%, White Mica: 86.26% and Pyrite: 84.75%. The average confidence level across all classes is 83.08%.
These results provide a clear indication of the model's performance on the test set, reflecting its confidence in classifying each mineral type. We hope this information satisfies your request and enhances the understanding of our model's performance. Thank you again for your constructive comments.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 7: More detailed explanation in the conclusion section.
RESPONE 7: We appreciate your valuable feedback. In response to your suggestion, we have revised the conclusion section to provide a more detailed explanation of our findings and their implications. Specifically, we have elaborated on the key contributions of our study, the significance of the model’s performance under controlled conditions, and the potential impact of addressing the identified limitations for practical applications. We have also highlighted the importance of future work in enhancing the model’s robustness and applicability in real-world scenarios. The updated conclusion section now offers a comprehensive summary that better reflects the depth of our research and the broader context of its contributions. [These revisions have been included in the paragraph of chapter 7 of the manuscript]
Thank you for your insightful comments, which have helped us to strengthen the conclusion of our manuscript.
Reviewer 2 Report
Comments and Suggestions for AuthorsThe manuscript describes and discusses a deep learning-based technology for the intelligent identification of various minerals. It is generally well written and well structured. Moreover, both the research objective (i.e., intelligent identification of minerals) and the conceptualized methodology (i.e., deep architectures adopted) are of scientific value and attraction.
Therefore, I believe that the manuscript satisfies the standards for publication in Minerals. However, several minor items still need to be addressed before the manuscript is accepted for publication.
Lines 22-27:
This sentence is too long and the meaning of the authors is unclear. Please rewrite.
Line 30:
mineral intelligent recognition → intelligent mineral recognition?
Line 76:
And → and?
Line 76:
has achieved → have achieved?
Line 102:
Application → application?
Line 275:
module outputs the dimensionality reduction features →
module outputs the dimensionally reduced features?
Line 281:
What do you mean by the first sentence? It seems to be an incomplete sentence! Please reconsider.
Line 283:
the Settings → the settings?
Line 326:
and Angle of objects → and angle of objects?
Line 333:
mineral intelligent recognition model → intelligent mineral recognition model?
Line 393:
Graph of the box_loss curves on the validation dataset and the test dataset. →
Graph of the box_loss curves on the validation dataset and the train dataset???
Line 409:
Graph of the cls_loss curves on the validation dataset and the test dataset. →
Graph of the cls_loss curves on the validation dataset and the train dataset???
Line 419:
Graph of the dfl_loss curves on the validation dataset and the test dataset. →
Graph of the dfl_loss curves on the validation dataset and the train dataset.
Line 505:
mineral intelligent recognition → intelligent mineral recognition?
Lines 512-518:
This paragraph is a verbatim copy of the previous paragraph (lines 505-511)!
Line 525:
mineral intelligent recognition → intelligent mineral recognition?
Comments on the Quality of English LanguageI recommend that the manuscript be partially revised in terms of English language to improve the quality of writing.
Author Response
COMMENT 1:
Lines 22-27: This sentence is too long and the meaning of the authors is unclear. Please rewrite.
Line 30: mineral intelligent recognition → intelligent mineral recognition?
Line 76: And → and?
Line 76: has achieved → have achieved?
Line 102: Application → application?
Line 275: module outputs the dimensionality reduction features →module outputs the dimensionally reduced features?
Line 281: What do you mean by the first sentence? It seems to be an incomplete sentence! Please reconsider.
Line 283: the Settings → the settings?
Line 326: and Angle of objects → and angle of objects?
Line 333: mineral intelligent recognition model → intelligent mineral recognition model?
Line 393: Graph of the box_loss curves on the validation dataset and the test dataset. →Graph of the box_loss curves on the validation dataset and the train dataset???
Line 409: Graph of the cls_loss curves on the validation dataset and the test dataset. → Graph of the cls_loss curves on the validation dataset and the train dataset???
Line 419: Graph of the dfl_loss curves on the validation dataset and the test dataset. → Graph of the dfl_loss curves on the validation dataset and the train dataset.
Line 505: mineral intelligent recognition → intelligent mineral recognition?
Lines 512-518: This paragraph is a verbatim copy of the previous paragraph (lines 505-511)!
RESPONE 1:
Dear Reviewer,
We would like to express our sincere gratitude for your thorough and careful review of our manuscript. Your detailed comments and suggestions have been invaluable in helping us improve the quality of our work.
In response to the issues you raised regarding formatting and expression, we have made the necessary corrections throughout the manuscript. We have also conducted a comprehensive review of the entire document to ensure consistency in formatting and clarity in expression.
We greatly appreciate the time and effort you invested in reviewing our paper, and we are confident that the revisions have significantly enhanced the quality of our work. Thank you once again for your insightful feedback and for helping us refine our manuscript.
Sincerely,
Luhao He
Reviewer 3 Report
Comments and Suggestions for AuthorsThe paper titled “Application of Target Detection based on Deep Learning in Intelligent Mineral Identification” discusses the application of deep learning techniques in the field of mineral identification. The authors have demonstrated the effectiveness of their model through various key metrics such as Precision, Recall, mAP50, and mAP50-95. While the paper presents significant findings and potential practical applications, it also identifies certain challenges, especially in cases involving multiple sample types where the model may misclassify due to similarities in morphology, color, or luster.
Below are the detailed issues and suggestions for improvement:
1、Although YOLOv8 is the latest general model in the YOLO series and its advantages are detailed in the article, it is still necessary to use evaluation metrics to make a simple comparison with previous models (such as YOLOv7) to highlight its advantages.
2、The paper mentions the use of various data enhancement strategies but does not provide detailed descriptions of these methods. It is recommended to include a comprehensive description of each data augmentation technique implemented and explain their impact on model performance.
3、Although the paper provides the model’s performance in terms of Precision, Recall, etc., it lacks an in-depth analysis of the results. It is suggested to add a detailed analysis of the experimental results, explaining the significance of each metric and its impact on mineral recognition.
4、The paper mainly focuses on testing under laboratory conditions and lacks discussion of practical application scenarios. How does the model perform in practical application scenarios, and is it possible to evaluate its applicability in real-world environments?
5、As shown in Table 1, the number of training samples for each category is not equal. Ideally, the number of training samples for each category should be consistent. How does this article address the issue of data imbalance? Or is the model minimally affected or unaffected by data imbalance?
6、Combining Figure 5 and the description in line 378, the training and validation losses peaked in the fourth epoch. However, the description mentions that the validation loss was 1.3143, while Figure 5 shows that the validation loss is less than 1.1, which is inconsistent.
7、Combining Figure 7 and the description of lines 416-417, the loss of the training set and the validation set are stable at 0.30692 and 0.90743 respectively. However, the figure does not show that the loss of the training set is stable at 0.30692.
8、Some paragraphs are not clearly expressed, and there are some minor issues with the format. It is recommended to carefully polish the language and proofread the format of the entire paper to ensure clear expression and rigorous logic.
Comments on the Quality of English LanguageSome paragraphs are not clearly expressed, and there are some minor issues with the format. It is recommended to carefully polish the language and proofread the format of the entire paper to ensure clear expression and rigorous logic.
Author Response
COMMENT 1: Although YOLOv8 is the latest general model in the YOLO series and its advantages are detailed in the article, it is still necessary to use evaluation metrics to make a simple comparison with previous models (such as YOLOv7) to highlight its advantages.
RESPONE 1: Thank you for your valuable feedback. We appreciate your suggestion to include a comparison with previous models to highlight the advantages of YOLOv8. In response to your comments, we have made the following revisions:
We have added a comparison of the YOLO series versions, specifically focusing on speed and accuracy metrics, in the Introduction section. This comparison includes YOLOv7 and YOLOv8, and it clearly illustrates the performance improvements in terms of mean Average Precision (mAP) and inference time (ms/img). By including this information, we aim to provide a clearer understanding of YOLOv8's advancements over its predecessors.
[These revisions have been included in the fifth major paragraph of chapter 1 of the manuscript]
We hope that this addition meets your expectations and strengthens the manuscript.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 2: The paper mentions the use of various data enhancement strategies but does not provide detailed descriptions of these methods. It is recommended to include a comprehensive description of each data augmentation technique implemented and explain their impact on model performance.
RESPONE 2: Thank you for your valuable feedback and for pointing out the need for more detailed descriptions of the data augmentation techniques used in our study. In response to your comments, we have added a comprehensive description of the data augmentation strategies implemented in our model. Specifically, the methods include:
1, Flipping: We applied horizontal and vertical flips to simulate variations in the orientation of the minerals. This helps the model learn from different directions, which is particularly useful for detecting minerals that may appear in various positions within the image.
2,Translation: Small random translations were used to slightly shift the mineral images in different directions. This augmentation helps the model better handle scenarios where minerals are not perfectly centered in the image, ensuring that the model is not overly sensitive to the exact position of the minerals. This, in turn, improves the model’s ability to detect and classify minerals under suboptimal conditions.
3,Rotation: Considering that minerals may have irregular shapes and orientations, we employed rotation augmentation to simulate different perspectives of the minerals. This technique enhances the model’s capability to detect and classify minerals from various angles, ensuring high accuracy even when the target objects are rotated in different directions.
[These revisions have been included in the paragraph of chapter 3.2 of the manuscript]
These augmentations have been shown to improve the model's robustness and overall performance. We believe that this additional explanation provides a clearer understanding of the role and impact of each data augmentation technique in our study.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 3: Although the paper provides the model’s performance in terms of Precision, Recall, etc., it lacks an in-depth analysis of the results. It is suggested to add a detailed analysis of the experimental results, explaining the significance of each metric and its impact on mineral recognition.
RESPONE 3: Thank you for your insightful feedback and for taking the time to review our manuscript. We appreciate your suggestion to include a more in-depth analysis of the experimental results, particularly in explaining the significance of each performance metric and its impact on mineral recognition.In response to your comments, we have made the following revisions to the manuscript:
We have added a detailed analysis of the model's performance across the seven minerals using Precision-Recall (PR) curves, as illustrated in the revised Figure 10. This analysis includes an explanation of the significance of the area under the PR curve (AUC-PR) and the [email protected] metric, providing a clearer understanding of how these metrics reflect the model's detection capabilities for each mineral. Additionally, we have interpreted the PR curves for each mineral based on their respective [email protected] values and discussed the implications of these results, particularly in relation to the model's precision and recall across different thresholds. [These revisions have been included in the seventh major paragraph of chapter 4.2 of the manuscript]
These revisions aim to provide a comprehensive understanding of the model's performance, addressing the points you raised. We believe that this additional analysis enhances the manuscript's depth and provides valuable insights into the model's strengths and areas for improvement. Once again, we sincerely thank you for your constructive feedback, which has been instrumental in improving the quality of our work.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT4: The paper mainly focuses on testing under laboratory conditions and lacks discussion of practical application scenarios. How does the model perform in practical application scenarios, and is it possible to evaluate its applicability in real-world environments?
RESPONE4: Thank you for your valuable feedback and for highlighting the need to discuss the model's performance in practical application scenarios. We appreciate your concern about the model's applicability in real-world environments, as this is a crucial aspect of our research.
In response to your comments, we have expanded the discussion section to address these points. We now acknowledge that while the deep learning model demonstrates high accuracy in mineral recognition tasks under controlled laboratory conditions, there are limitations when it comes to real-world applications. Specifically, the model has shown instances of misclassification, particularly with samples containing multiple minerals, as demonstrated in Figure 12.
To provide a more comprehensive evaluation of the model's potential in practical settings, we have proposed several strategies for future research. These include increasing the diversity of training samples, optimizing the model architecture, and incorporating post-processing techniques to reduce misclassification. Additionally, we emphasize the importance of testing the model in real-world environments, such as active mining sites and geological survey locations, to assess its stability, effectiveness, and reliability under varying environmental conditions.
[These revisions have been included in the paragraph of chapter 6 of the manuscript]
We believe that these discussions and proposed future directions will help to better address the concerns raised and provide a clearer pathway for evaluating the model's applicability in real-world scenarios. Once again, we sincerely thank you for your thorough review and constructive suggestions, which have significantly contributed to the improvement of our manuscript.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 5: As shown in Table 1, the number of training samples for each category is not equal. Ideally, the number of training samples for each category should be consistent. How does this article address the issue of data imbalance? Or is the model minimally affected or unaffected by data imbalance?
RESPONE 5: Thank you for your valuable feedback regarding the issue of data imbalance in our study. We appreciate your attention to this important aspect. In response to your comments, we have taken the following steps to address the data imbalance issue:
1, Addressing Data Imbalance: While the number of training samples per category is indeed unequal, we have explored several strategies to mitigate the effects of this imbalance. Specifically, we have experimented with various loss functions designed to handle imbalanced datasets, such as the Weighted box_Loss , cls_Loss and dfl Loss. These loss functions help adjust the model’s focus towards underrepresented classes and improve overall performance.
2, Image Augmentation: To further counteract data imbalance, we applied various image augmentation techniques such as flipping, translation, and rotation. These augmentations help to artificially increase the diversity of training samples for underrepresented classes, thereby providing more balanced input data.
Model Performance and Data Imbalance: Despite the inherent data imbalance, our model has demonstrated satisfactory performance across the different mineral categories. In our revised discussion section, we have included an analysis of how the model’s performance is affected by the data imbalance. We highlight areas where the model performs well and acknowledge the limitations where the imbalance has a noticeable impact. Future Directions: We have also suggested potential avenues for future work to further address data imbalance. This includes exploring data augmentation techniques specific to minority classes and employing advanced sampling methods to create a more balanced training dataset.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 6: Combining Figure 5 and the description in line 378, the training and validation losses peaked in the fourth epoch. However, the description mentions that the validation loss was 1.3143, while Figure 5 shows that the validation loss is less than 1.1, which is inconsistent.
RESPONE 6: Thank you for your meticulous review and for pointing out the inconsistency regarding the validation loss. We appreciate your attention to detail. Upon re-evaluating the data, we discovered that the validation loss was indeed reported incorrectly. We have corrected the validation loss value from 1.3143 to the accurate value of 0.71715. This adjustment aligns with the validation loss depicted in Figure 5. We have thoroughly reviewed and cross-checked all related data to ensure that there are no further discrepancies. The manuscript has been updated accordingly to reflect these corrections. Thank you once again for your careful examination and valuable feedback, which has helped us ensure the accuracy of our results.
——————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT 7: Combining Figure 7 and the description of lines 416-417, the loss of the training set and the validation set are stable at 0.30692 and 0.90743 respectively. However, the figure does not show that the loss of the training set is stable at 0.30692.
RESPONE 7: Thank you for your careful review and for highlighting the discrepancy regarding the stability of the training set loss. We appreciate your attention to detail. Upon re-evaluating the data, we found that the reported training set loss of 0.30692 was incorrect. The correct value should be 0.96787, which is now consistent with the observations presented in Figure 7. We have updated the manuscript to reflect this corrected value and have thoroughly reviewed all related data to ensure accuracy. The revised figure and description now correctly represent the stability of the training set and validation set losses. Thank you once again for your diligent examination, which has been invaluable in improving the precision of our work.
———————————————————————————————————————————————————————————————————————————————————————————————————————
COMMENT8: Some paragraphs are not clearly expressed, and there are some minor issues with the format. It is recommended to carefully polish the language and proofread the format of the entire paper to ensure clear expression and rigorous logic.
RESPONE 8: Thank you for your insightful feedback and for highlighting the areas that need improvement in our manuscript. We appreciate your recommendations regarding the clarity of expression and formatting issues.
In response to your comments, we have thoroughly revised and polished the language throughout the paper to ensure clearer expression and more rigorous logic. Additionally, we have carefully proofread the entire manuscript to correct any formatting inconsistencies and enhance overall readability.
We believe these revisions have significantly improved the clarity and presentation of the paper. Thank you once again for your valuable feedback, which has greatly contributed to enhancing the quality of our work.
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsNo further comments
Comments on the Quality of English LanguageNo comments
