Machine Learning-Based Label Quality Assurance for Object Detection Projects in Requirements Engineering

Round 1

Reviewer 1 Report

Dear Authors,

Thank you for submitting your work to the Applied Sciences-MDPI journal. Some improvements need to be considered to make your work better.

Comments:

Featured Application: links are not working!

Abstract:

-        Please add two sentences about experiments you have done and the results you have reached.

1.      Introduction

-        At the end of this section, I suggest you add a paragraph representing the article map which briefly describes what each section will discuss.

2.      Related Work

-         I suggest adding a brief paragraph describing what this section will discuss.

 - Related work is a bit short try to expand it and add the most recent works (i.e. from 2023), please.

3. Proposed Process for Data Requirements within Requirements Engineering

- Some abbreviations, such as MRI in this section, should be defined the first time mentioned.

5. Results and Discussions 

- Please transfer the paragraph starting with “….This approach can be extended to other tasks beyond detection. For instance, for a  segmentation task, polygons could be used instead of bounding boxes. The problem solver is responsible for determining the representation of the label and uncertainty region ranges” to be in conclusion as a future work for your framework.

References:

Please revise the references and add some of the missing conferences’ locations such as “Ramesh, A., Pavlov, M., Goh, G., Gray, S., Voss, C., Radford, A., Chen, M. and Sutskever, I., 2021, July. Zero-shot text-to-image  generation. In International Conference on Machine Learning (pp. 8821-8831). PMLR. And  “4. Rombach, R., Blattmann, A., Lorenz, D., Esser, P. and Ommer, B., 2022. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 10684-10695).”

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Dear Authors,

the article is relevant and scientific-sounding; however, there are some notes that should be revised before promotion:

1. There is no summary of the findings and provisions in Results section. 

2. Along with this, there is no Discussion section. So all the advances of Authors' findings, as well as limitations of the study and future perspectives are not clear. 

3. The References list should be expanded with new internationsl literature. 

4. Conclusion section should be expanded and given in details.

Good luck!

Dear Authors,

please carefully check the spelling.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Submitted manuscript need several modifications and my comments are as follows:

1. Title seems to be very general.Further it is unclear what exactly authors want to convey to scientific readers.

2. How label quality assurance for object detection is achieved ? Any quantitative or qualitative parameter.

3. Since the work of authors specifically focused on images, an important aspect related to filter/denoise of images using Wavelet, GLCM is left unaddressed. Kindly add a specific paragraph related to the influence of above-mentioned terminology after referring following journals:

a. https://ieeexplore.ieee.org/abstract/document/8038631

b. https://link.springer.com/article/10.1134/S0361768820080113

c. https://www.mdpi.com/1424-8220/22/18/7010

4. Authors used NN machine learning model as evident from Fig.5.It is always desirable to compare the results with different ML models.It is strongly recommended to include comparison results in revised version.For more clarity,refer 3a.

5. There are so many small paragraphs. Kindly re-organized the whole structure of manuscript.

6. The description added in conclusion section is very unclear. The statements written must be supported by certain key findings, qualitative and quantitative results etc.

7. How authors proposed paper is different than already published literature.It should be specifically mentioned and added in revised version.

8. Kindly stick with single reference style which is uniform across all references.

Language is ok.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Dear Authors,

Thank you for re-submitting your work to the Applied Sciences journal. All my comments and suggestions have been covered successfully.

Thank you

Reviewer 2 Report

Good luck!

Reviewer 3 Report

Manuscript has been revised based on suggestions.

English is fine.