An Efficient Approach for Detecting Driver Drowsiness Based on Deep Learning
Round 1
Reviewer 1 Report
This paper considers detecting driver drowsiness problems using deep learning. This is one of the widely investigated problems for self-driving. For this, the authors proposed two approaches with several scenarios. However, the current version of the manuscript has many parts that need to be modified:
- The proposed algorithm should be introduced in the abstract.
- It is necessary to introduce more related papers dealing with the issues in the introduction section.
- It would be better to give a figure in the introduction that explains the problem.
- Regarding the use of figures with citations, if there is no permission from the authors, it should be redrawn.
- Some texts used in the figures (ex, Figure 5) are not clearly visible.
- It would be better to modify the algorithm pseudocode in the section 3.1.4 to make it easier to see.
- There is no detailed explanation and description for the method 2 and figure 9 in section 3.2.
- The performance of the proposed algorithm should be compared with other base algorithms. However, there is no base algorithm in this manuscript.
- It would be better to cite more recent papers related to the problem and to compare the result with the proposed methods.
Author Response
Dear Reviewer,
Thank you very much for your comments concerning our manuscript titled: “An Efficient Approach for Detecting Driver Drowsiness Based on Deep Learning” (ID: applsci-1349479). Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our research. We have carefully taken the comments into account and have made corrections, which we hope will meet with approval. Our responses to your comments are described below in a point-to-point manner. Appropriated changes, suggested by you, have been introduced to the revised manuscript (highlighted within the document). The main corrections in the revised manuscript and the responses to your comments are in the attachment.
Please see the attachment.
We appreciate the very positive and constructive comments from you. We concur with your comments and suggestions that help us to have the corrections to meet your approval.
Once again, thank you very much for your precious time and consideration. We look forward to hearing from you at your earliest convenience.
Yours sincerely,
Anh-Cang PHAN
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Dear authors, the article is suitable for publication after the comments have been eliminated.
1. The list of references needs to be updated, it is too small.
2. Conclusions are too scanty and do not reflect the whole essence of your research.
3. The abstract is not informative.
4. Drawings are poor presentation.
5. The number of participants in the experiment is too small and there is no way to judge the quality of experimental studies.
Author Response
Dear Reviewer,
Thank you very much for your comments concerning our manuscript titled: “An Efficient Approach for Detecting Driver Drowsiness Based on Deep Learning” (ID: applsci-1349479). Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our research. We have carefully taken the comments into account and have made corrections, which we hope will meet with approval. Our responses to your comments are described below in a point-to-point manner. Appropriated changes, suggested by you, have been introduced to the revised manuscript (highlighted within the document). The main corrections in the revised manuscript and the responses to your comments are in the attachment.
Please see the attachment.
We appreciate the very positive and constructive comments from you. We concur with your comments and suggestions that help us to have the corrections to meet your approval.
Once again, thank you very much for your precious time and consideration. We look forward to hearing from you at your earliest convenience.
Yours sincerely,
Anh-Cang PHAN
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Many concerns have been addressed in this revised version. However, some questions still remain and some parts need to be corrected.
- The proposed methods seem very similar to the prior works [12] for the facial landmark and [13] deep learning approach. The authors should clearly describe how the proposed methods differ from the prior methods in the manuscript.
- Further, in Table 9, some approaches of related works also achieve 96 - 97.1 % accuracy for the problem. Then, what is the main advantage of the suggested approach of the manuscript compared to them? It should be explained more clearly.
- Most of the texts in the figures are not clearly visible.
- The terms "formula" should be changed to "equation" in the manuscript.
- In this pseudocode of Algorithm1, "all" and "each" are expressed as duplicates in the "for" statement.
Author Response
Dear Reviewer,
Thank you very much for your comments concerning our manuscript titled: “An Efficient Approach for Detecting Driver Drowsiness Based on Deep Learning” (ID: applsci-1349479). Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our research. We have carefully taken the comments into account and have made corrections, which we hope will meet with approval. Our responses to your comments are described below in a point-to-point manner. Appropriated changes, suggested by you, have been introduced to the revised manuscript (highlighted within the document).
Please see the attachment.
Once again, thank you very much for your precious time and consideration. We look forward to hearing from you at your earliest convenience.
Yours sincerely,
Anh-Cang PHAN
Author Response File: Author Response.pdf
Reviewer 2 Report
Dear Authors. Well done, you corrected and supplemented the article well.
Round 3
Reviewer 1 Report
Most of my concerns have been addressed.
