Research on the Clothing Classification of the She Ethnic Group in Different Regions Based on FPA-CNN

Round 1

Reviewer 1 Report

The manuscript entitled presented clothing classification of 5 different regions of China. The paper is written nicely, and supported by clear presentation of experimental setup and results. Although the work seems interesting and within the scope of the journal, a few corrections are suggested:

1. The phrase "She Ethnic Groups" seems confusing. If possible, use alternate one.

2. Introduction is nice, but can be strengthened by including major contributions (bulleted list), and a paper organization (at the end of the section).

3. The visibility of some figures (Fig. 1, 5(b)) should be improved.

4. It is mentioned in the paper that "She ethnic clothing from different regions has similarities in shape and texture features, but their color features have a certain degree of discriminability". Based on this concept, the color is considered as one of the main components to classify the regions. But in case, the same color is adopted by different regions, then how the system will classify the regions. Please justify the answer and include the same in manuscript.

5. Layers of the CNN should be presented with the help of a figure.

6. Few typos:

(a) Line 258: "CNN network" should be corrected to "CNN"

(b) Line 259: x must be replaced with a multiply symbol.

(c) In Section 3.2.2, it is mentioned that “From the table, it can be observed that”. Please mention table number.

(d) Please proofread the whole paper again and correct such typos (if present).

7. The pseudo-codes of PSO and FPA should be included in the paper for better understanding. A reason to justify "Why PSO and FPA" may be included.

8. Axis labels to Fig 7 (a) - (e) should be provided.

9. The convergence graph in Fig. 5(a) provides a clear understanding on the performance of PSO and FPA. Conclusions seems novel and supported by results.

10. No references from the target journal. Try to include some references from the target journal so that the submission seems appropriate.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

This paper presents a method for image classification based on CNN and color features. The paper is easy to read and has a good structure. Here are some points that need to improve in the future.

- The authors should explain  why only a few color space is considered ?

- The feature fusion can be applied before or after feature extraction as we can combine color channel ?

- The dataset is so small for an approach based CNN. Why don't you collect more images or using data augmentation methods ? 

- Some potential references are related to your papers and should be cited within the text:

https://doi.org/10.1117/1.JEI.27.1.011010

https://doi.org/10.3390/s22072660

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Some changes are needed to raise the level of study. My suggestions are listed below:

1) The introduction section is insufficient. It should be expanded. In particular, what is the novelty of the study? What is his contribution to the literature? It should explained in detail in the introduction.

2) Why FPA was used in this study. How is it different from other meta-heuristics?

3) You can benefit from the following studies on ANN and FPA.

- Kaya, E., & Baştemur Kaya, C. (2021). A novel neural network training algorithm for the identification of nonlinear static systems: Artificial bee colony algorithm based on effective scout bee stage. Symmetry, 13(3), 419.

- Kaya, E. (2022). Quick flower pollination algorithm (QFPA) and its performance on neural network training. Soft Computing, 26(18), 9729-9750.

- Kaya, C. B., & Ebubekir, K. A. Y. A. (2021). A novel approach based to neural network and flower pollination algorithm to predict number of COVID-19 cases. Balkan Journal of Electrical and Computer Engineering, 9(4), 327-336.

4) FPA was compared to PSO only. This is not enough. For comparison, I suggest adding 5 or 6 more recent meta-heuristic algorithms. In this way, you can better demonstrate the success of the proposed method.

5) You should create a discussion section.

Author Response

Dear Reviewer:

      Thank you very much for your constructive comments feedback. we have revised the paper as carefully as was possible in the hopes of making it more complete. We are happy to receive additional feedback and suggestions.

     Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The authors have addressed all of my comments. I am satisfied with the revised version. I believe this article is ready for publication.

Reviewer 3 Report

The submission should be accepted.