All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
Dear authors,
Thank you for revising the manuscript. Reviewers did not respond to the invitation to review the revision. When I examined your article, it is thought that the necessary additions and arrangements have been performed according to the editor and reviewers' opinions and the article has been sufficiently improved. As such, the article is considered acceptable. Furthermore, when submitting the final file, attention is needed for correct sentence formation including equations as they are part of the sentences. Equations should be used with correct equation number. Please do not use “as follows”, “given as”, etc. All variables should be written in italic as in the equations. Their definitions and boundaries should be defined. Only one character must be used for the multiplication operator. Different characters should not be mixed for multiplication.
Best wishes,
[# PeerJ Staff Note - this decision was reviewed and approved by Xiangjie Kong, a PeerJ Section Editor covering this Section #]
Dear authors,
Thank you for submitting your article. Reviewers have now commented on your article and suggest major revisions. When submitting the revised version of your article, it will be better to address the following:
1. The abstract does not present the creation or usage of the dataset.
2. The problem statement is not clear in the Introduction. This section should include a problem definition, motivation, overview of the proposed solution, enumeration of the contributions and organization of the paper. Organization of the paper should be given at the end of Introduction section.
3. Evaluation of the techniques covered in the literature is missing. Negative and positive aspects of these techniques should be stated. The lessons learned from the literature is absent. These are the key aspects to motivate for the research and for new researchers who want to tackle the same problem.
4. All reported graphs should be accompanied by some concrete description of the lessons learned from the results reflected in the graph. It is important to explain them in detail and to enrich them with some semantics by showing the reasons for these results, how they can be further improved, etc.
5. In the conclusions, please state explicitly what lessons can be learnt from this study and then describe in more detail the future research directions.
6. English grammar and writing style errors should be corrected. There are many writing errors that should be corrected.
[# PeerJ Staff Note: The review process has identified that the English language must be improved. PeerJ can provide language editing services if you wish - please contact us at [email protected] for pricing (be sure to provide your manuscript number and title). Your revision deadline is always extended while you undergo language editing. #]
The manuscript requires formatting adjustments to adhere to the journal's guidelines, particularly regarding the citation of references. Moreover, the English language should be carefully revised to enhance readability, ensuring that native English speakers can comprehend the text without ambiguity.
The manuscript presents a thorough effort in advancing image segmentation for transmission line foreign object detection, yet lacks clear explanation on the transformative impact of the introduced module enhancements. Clarification is needed on the specific advantages this modified RB-UNet offers over directly employing the latest segmentation models, especially regarding its unique contributions to real-world applicability.
The generalization capability of the proposed method is a paramount concern, especially considering the need for consistent performance across a wide range of foreign object types and variations found under diverse real-world conditions, To gauge the method’s generalization, assessing performance with broader intra-class differences is important. Given the practical context of varied image sources in transmission line inspections, a detailed analysis of the model's applicability across real-world operational scenarios, beyond the internet-derived dataset, would strengthen the study's validity.
Overall, the study was conceived and implemented good, but there are some areas that need to be improved before publication, especially in terms of language clarity, dataset description, and innovative statements.
-This paper proposes a foreign object segmentation method based on the RB-UNet algorithm; however, the motivation behind the design of this algorithm is not clearly articulated. Previous works often suffer from a lack of identification accuracy and fail to address the issue of class imbalance effectively, presenting it as an overly broad problem, which does not clearly define a research gap. The paper needs to explicitly state the research aim or research question in the introduction.
-The motivation presented in the article is limited. The contribution section resembles a technical report rather than an academic paper. For example, incorporating ResNet50 as a feature extractor should not be considered a part of the contributions, as it merely represents a non-constructive network concatenation operation. It is essential to focus on describing the correlation between this module and your research question within the motivation section of the introduction. Given that the research question itself lacks clarity, there is a need to rewrite the introduction to ensure that there is a coherent logical connection between the introduction and the motivation of the paper.
-The layout of the paper requires improvement.
- The resolution of the images needs to be enhanced, as the clarity of Figure 5 is currently insufficient and blurred. It is recommended to import images into the document in the form of vector graphics rather than using screenshots.
-The ablation experiments in the experimental section are comprehensive, which is commendable. However, there are too few benchmark experiments. The authors should add more horizontal comparisons with baseline models, rather than solely focusing on the vertical comparative experiments of their own model.
-It is recommended to revise Figure 1 and Figure 2, especially Figure 1. The UNet network structure is a common backbone and does not require detailed depiction in text. The core design related to Figure 2 should emphasize the original structures of your work, rather than merely illustrating simple information flow.
-Please avoid using underscores (_) for subscripts in the methodology section. This format detracts from the readability of the paper.
-The writing in the methodology section is isolated in logic. It needs to resonate textually with the research questions, objectives, and motivations. Enhancing performance should not be presented as the core objective of network optimization. The network design in the paper must reflect a response to specific research questions.
-In Figure 5, the visual distinctions between each compared baseline are almost indiscernible. The network's contribution to improving segmentation results appears to be limited.
-The ablation study section should also include visual comparison charts. The impact of using different loss functions and swapping various modules on segmentation performance should be presented in a visual format to clearly demonstrate their effects.
The overall quality of this paper needs to be enhanced. This involves clarifying the research motivations, expanding on the experimental comparisons with baseline models, improving the integration of methodology with research objectives, enhancing the visual clarity of figures, and more clearly highlighting the novel aspects of the proposed network design. Each of these adjustments will contribute to a more robust and persuasive academic presentation.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.