Visual Lip Reading Dataset in Turkish

Round 1

Reviewer 1 Report

The originality of this study is non-syntheticity of the dataset.
Was this study approved in any ethical review board?
Did you get a written informed consent from each of subjects, for example, as shown in Figs.4,
6 and 7?
There is no statistical evidence for the normal distribution and right-skewed distribution.
Minor corrections are as follows:
・”Defense industry” is unfavorite and is not described in 4.Conclusion.
・ at page 1, in 1.Summary, the second paragraph, last line: “these sentences” is to be
eliminated.
・at page 2, the second paragraph, the 3rd and 11th lines: Atila and Sabaz⇒Atila and Sabaz
[1]; the 3rd paragraph, the 1st line: “they” is to be eliminated; the 5th line: that⇒where; the 4th
paragraph, the 12th line: the Hide Markov Model⇒the hidden Markov model.
・at page 3, the 2nd paragraph, the 6th line: Binyan is to be eliminated.
・at page 4: Figure 1 and Table 1 are equivalent. So, either is to eliminated.
・at page 6, Section 3.2, the 2nd line: the popular python library OpenCV⇒the popular Python
library, OpenCV; what does “a function” mean?
・at page 7, Figure 5: two-digit numbering cannot be perceived.
・at page 8, 4.Conclusion, the 2nd line: the [1] dataset⇒the dataset [1]

Comments for author File: Comments.pdf

Author Response

We would like to thank you for reviewing our "Visual Lip Reading Dataset in Turkish" manuscript that we submitted to MDPI Data Journal and for contributing to the development with your comments. Details of the revisions in the article and our responses to the comments of the referees are attached.

Author Response File: Author Response.docx

Reviewer 2 Report

(1) All these data are from Youtube. If the face data are released, will it violate some rules?  Is it legal to public and process the face data on the website?

(2) Besides the English dataset, there are also some other lipreading datasets in other languages, e.g. LRW-1000 (Mandarin), LRWR(Russian), and these data should also be introduced in the related work part, to avoid giving the impression that all the existing lip reading datasets are in English. 

Author Response

Before explaining the revisions, we would like to thank you for reviewing our "Visual Lip Reading Dataset in Turkish" manuscript that we submitted to MDPI Data Journal and for contributing to the development with your comments. Details of the revisions in the article and our responses to the comments of the referees are explained below, point by point.
According to reviewer 2 we applied the revisions in terms of;
⦁    “All these data are from YouTube. If the face data are released, will it violate some rules?  Is it legal to public and process the face data on the website?”  
We found the statement about the legal principle of fair use, which indicates that an open-source content is not republished for the same purpose, and copyright is not sought in cases where it is used for educational purposes in YouTube’s content usage policies. Furthermore, Sharma et al. [1] used the videos they obtained directly, and their data article was accepted as "Not applicable" in the Informed Consent Statement section as we updated the same statement similarly.
⦁    “Besides the English dataset, there are also some other lipreading datasets in other languages, e.g., LRW-1000 (Mandarin), LRWR(Russian), and these data should also be introduced in the related work part, to avoid giving the impression that all the existing lip-reading datasets are in English.”
LRW-1000 [2] and LRWR [3] studies were reviewed in the literature section. In this way, lip-reading studies in Russian and Mandarin languages have been included in the literature review.
In addition to these revisions, the article was re-examined by all authors and English improvements were made throughout the article. The references were checked again, paying attention to their compatibility with the contents of the manuscript.
We confirm that neither the manuscript nor any parts of its content are currently under consideration or published in another journal.
All authors have approved the manuscript and agree with its submission to MDPI Data Journal.
We have no conflicts of interest to disclose. Please address all correspondence concerning this manuscript to us via our emails.
Thank you for your consideration of this manuscript.

Author Response File: Author Response.docx