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.
The paper is well written and well presented. It provides good background information on banana leaf disease detection. I am happy to accept the paper.
[# PeerJ Staff Note - this decision was reviewed and approved by Xiangjie Kong, a PeerJ Section Editor covering this Section #]
well-structured, concise, and informative. It provides sufficient background on the importance of banana leaf disease detection and clearly states the scope of the review. The language is scientific, but accessible, and the abstract follows a logical flow from problem statement → methodologies → challenges → contributions. Minor refinement could further improve clarity (e.g., sentence at line 29 could be smoothed for readability).
The study is designed as a systematic review, focusing on Artificial Intelligence-based techniques for banana leaf disease detection. It highlights methods such as image preprocessing, machine learning, deep learning, and transfer learning, while paying special attention to lightweight architectures suitable for real-world agricultural applications. The design also includes discussion on datasets, evaluation metrics, and a brief case study, which strengthens the comprehensiveness of the review.
the outlines valid insights by synthesizing multiple approaches and identifying both strengths and limitations of current AI techniques. The emphasis on lightweight deep learning models for resource-constrained environments enhances practical relevance. While findings seem credible and valuable, the extent of their validity depends on the range and quality of reviewed literature, which is not detailed in the abstract.
he inclusion of cultivar diversity and morphological variations is a strong point, as it emphasizes real-world challenges.
A brief mention of future directions (e.g., integration of multimodal data, real-time deployment) could further strengthen the abstract.
The paper addresses an interesting topic. And overall, it does a decent job. However, both the reviewers have identified areas that can be improved.
Hence, the paper needs careful revision while addressing the comments of the reviewers.
**PeerJ Staff Note:** Please ensure that all review and editorial comments are addressed in a response letter and that any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.
**Language Note:** The review process has identified that the English language must be improved. PeerJ can provide language editing services - please contact us at [email protected] for pricing (be sure to provide your manuscript number and title). Alternatively, you should make your own arrangements to improve the language quality and provide details in your response letter. – PeerJ Staff
The manuscript is written in generally professional and understandable English. However, several sentences are overly long, and some wording (e.g., “descry” instead of “describe,”). The manuscript should undergo thorough professional English editing to ensure clarity, eliminate grammatical inconsistencies, and conform to academic writing standards. The article covers a large number of studies and technologies related to banana leaf disease detection. It references a rich selection of sources (from 2019 to 2025) The structure generally aligns with a systematic literature review. The inclusion of an abstract, introduction, methodology, and outlined research questions is appropriate. The flow is logical.
The article clearly falls within the Aims and Scope of PeerJ Computer Science, as it focuses on the application of Artificial Intelligence (AI), Machine Learning (ML), Deep Learning (DL), and Transfer Learning (TL) to detect diseases in banana leaves. The article type is explicitly a literature review, and the content is consistent with this classification.
As a review article, no new experimental work or direct ethical approval is required. However, technical rigor in the review process is expected. The paper includes a large number of cited studies and breaks them down by category (e.g., DL, ML, image processing), but the following issues affect perceived rigor:
* There is no detailed protocol (e.g., PRISMA or similar framework) describing how studies were identified, screened, and selected.
* There is no quality assessment (e.g., scoring or grading studies based on reliability, dataset quality, or reproducibility).
* Article selection appears to be based only on keyword search and recency, which may introduce bias.
Recommendation: The authors should describe a formal and repeatable review protocol, ideally including:
* Inclusion/exclusion criteria
* Duplicate removal method
* Quality scoring or ranking mechanism (if used)
The authors mention that articles were collected from Scopus, Google Scholar, IEEE, Web of Science, and ScienceDirect, using keywords such as: "banana leaf disease detection using DL", "AI-based banana diagnosis", etc.
While this is a good start, the methodology remains insufficiently detailed for full reproducibility:
* It is unclear how many articles per database were returned and selected.
* There is no description of how duplicates or irrelevant articles were excluded.
* The authors mention collecting ~150 articles, but do not provide a list or supplemental file with metadata (title, year, method, etc.).
Recommendation: Include a systematic review flow diagram (e.g., similar to PRISMA), and share a supplemental spreadsheet listing all included articles.
The review includes a very large number of AI models (CNNs, GANs, MobileNet, ResNet, SVM, federated learning, YOLO, etc.), but there is a strong imbalance toward certain types of studies:
* The review is overwhelmingly descriptive, summarizing dozens of models without comparing them or discussing their limitations.
* There is little mention of failure cases, negative results, or dataset limitations, which raises concerns about selection bias (only successful models reported).
* Real-world deployment aspects (e.g., inference time, memory footprint, usability for smallholder farmers) are mentioned only in passing.
Recommendation: Improve balance by:
* Discussing studies with lower accuracy or mixed results.
* Highlighting dataset weaknesses (e.g., small sample sizes, limited class diversity).
* Synthesizing performance into comparative tables with context (e.g., method used, dataset size, accuracy, F1 score).
As this is a literature review, PeerJ does not expect assessment of impact or novelty. The authors do not make unsupported claims of methodological innovation, which is appropriate.
The manuscript is generally well-written, with clear articulation of objectives and logical flow. The relevance of banana leaf disease detection to global agricultural productivity is well-established. However, the article would benefit from more structured reporting of the review methodology (e.g., databases searched, inclusion/exclusion criteria, number of papers reviewed). Providing tables or figures summarizing the reviewed techniques would improve clarity and transparency.
The paper is framed as a review, and it successfully identifies key AI techniques applied in the context of banana leaf disease detection. While the authors discuss machine learning, deep learning, and lightweight models effectively, the study lacks a clear description of the systematic review process. Incorporating a formalized framework (e.g., PRISMA flowchart) would enhance the robustness and reproducibility of the design.
The discussion on lightweight deep learning models, challenges related to cultivar variation, and the emphasis on real-world deployability is thoughtful and relevant. However, without a comparative summary (e.g., tabulated model performance or meta-analysis), the findings are largely qualitative. To enhance validity, quantitative insights and evidence-based comparisons should be included wherever possible.
lease consider detailing the review methodology, including databases searched and study selection criteria.
Adding a comparative table of different AI models, along with performance metrics, would greatly strengthen the utility of the review.
Including brief case studies or deployment examples could highlight practical relevance.
A discussion on explainability and model interpretability (e.g., use of Grad-CAM, SHAP) would add value given the practical implications in agriculture.
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.