Card3DFace—An Application to Enhance 3D Visual Validation in ID Cards and Travel Documents

Round 1

Reviewer 1 Report

The paper deals with an interesting topic face analysis with a specific focus on 3D. In the present form the paper needs improvements. Starting from the introduction authors should provide a wider picture on the research domain in order to better contestualise the proposed methodology. Some more references should be added as for instance the following ones:

DAGNES, Nicole, et al. Occlusion detection and restoration techniques for 3D face recognition: a literature review. Machine Vision and Applications, 2018, 29.5: 789-813.

WANG, Zhenzhou, et al. A New Face Recognition Method for Intelligent Security. Applied Sciences, 2020, 10.3: 852.

NONIS, Francesca, et al. 3D Approaches and challenges in facial expression recognition algorithms—a literature review. Applied Sciences, 2019, 9.18: 3904.

Also the methodological section needs improvements. More details should be provided regarding the overall methodology and the jey correlation in the specific method stages. Regarding the experimental validation more data regarding the experimental setting should be provided in order to have amore reliable perspective of the obatined performances 

Author Response

Coimbra, 30th of August 2021

Dear Sir

The authors would like to thank the reviewers and their comprehensive comments that definitely contributed to achieve a much better version of the submitted article.

We addressed all the comments in the revision and submitted a version of the paper in track changes (using Microsoft Word tools) and the PDF paper where we highlighted the most important changes applied to the paper. We also gone through the paper with an extensive English revision.

In the highlighted PDF version of the paper, we highlighted in yellow new paragraphs and in blue moved paragraphs.

We hope that this version will be considered by the reviewers as publishable.

The detailed discussion of the reviewer comments can be read bellow.

REVIEWER 1:

“The paper deals with an interesting topic face analysis with a specific focus on 3D. In the present form the paper needs improvements. Starting from the introduction authors should provide a wider picture on the research domain in order to better contestualise the proposed methodology. “

We reformulated sections 1, 2 and 3 with more contextualization paragraphs and a better separation between the section 2, stating the Related Work, and section 3, with the methodology.

“Some more references should be added as for instance the following ones:

 DAGNES, Nicole, et al. Occlusion detection and restoration techniques for 3D face recognition: a literature review. Machine Vision and Applications, 2018, 29.5: 789-813.

 WANG, Zhenzhou, et al. A New Face Recognition Method for Intelligent Security. Applied Sciences, 2020, 10.3: 852.

 NONIS, Francesca, et al. 3D Approaches and challenges in facial expression recognition algorithms—a literature review. Applied Sciences, 2019, 9.18: 3904.”

We also added some references missing, including these references kindly suggested by the reviewer. The PDF version of the paper highlights the most extensive changes and the Microsoft Word .docx file has all the changes made.

“Also the methodological section needs improvements. More details should be provided regarding the overall methodology and the jey correlation in the specific method stages.”

We provided additional insights on the methodologic section, mainly by creating a new section (section 3) with the methodology clearly separated from the Related Work. The PDF version of the paper highlights the most extensive changes and the Microsoft Word .docx file has all the changes made.

“Regarding the experimental validation more data regarding the experimental setting should be provided in order to have amore reliable perspective of the obatined performances” 

We provided more details in the experimental section. The PDF version of the paper highlights the most extensive changes and the Microsoft Word .docx file has all the changes made.

End of revision comments

One more time, thanks to the reviewers for their work.

With our best regards,

Leandro Dihl

Leandro Cruz

Nuno Gonçalves.

Author Response File: Author Response.pdf

Reviewer 2 Report

An end-user application to synthesize different views for lenticular cards is presented in this article 

• Extensive language editing shall be considered to improve the readability.
• The abstract seems to be deficient.
• The work presented in this article is very much similar to the article cited [16], apart from presenting this as a c++ and Qt-based application.
• There seems to be no significant contribution or value addition in the current version of the article. (Though there is a self citation)
• Restructuring the article shall be considered to highlight the novel contribution and advancements made clearly. 

Author Response

Coimbra, 30th of August 2021

Dear Sir

The authors would like to thank the reviewers and their comprehensive comments that definitely contributed to achieve a much better version of the submitted article.

We addressed all the comments in the revision and submitted a version of the paper in track changes (using Microsoft Word tools) and the PDF paper where we highlighted the most important changes applied to the paper. We also gone through the paper with an extensive English revision.

In the highlighted PDF version of the paper, we highlighted in yellow new paragraphs and in blue moved paragraphs.

We hope that this version will be considered by the reviewers as publishable.

The detailed discussion of the reviewer comments can be read bellow.

REVIEWER 2:

“An end-user application to synthesize different views for lenticular cards is presented in this article 

 Extensive language editing shall be considered to improve the readability.”

We make extensive revision in terms of English language. The revision can be analyzed in the Microsoft Word .docx file in track changes.

“The abstract seems to be deficient.”

We reformulated the abstract and added several details and insights on the paper that were missing.

“The work presented in this article is very much similar to the article cited [16], apart from presenting this as a c++ and Qt-based application.

There seems to be no significant contribution or value addition in the current version of the article. (Though there is a self citation)

Restructuring the article shall be considered to highlight the novel contribution and advancements made clearly. “

The paper in the citation [16] is a previous work of the authors, focused on the filtering method, allowing to achieve better tridimensional models of the head/faces. The method published in [16] is one specific part of the system pipeline described in the current article. The current article intends to be a system article where the whole end-to-end system is presented and demonstrated. We introduced this explanation on the text and highlighted in the PDF file to better visualization. The detailed changes are tracked in the Microsoft Word .docx file.

End of revision comments

One more time, thanks to the reviewers for their work.

With our best regards,

Leandro Dihl

Leandro Cruz

Nuno Gonçalves.

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper presents a system to enhance the ID card validation using 3D visual information. An image is captured and several viewpoint rendering is generated for face recognition. The system implementation pipeline is described and experimental results are demonstrated. 

Although the technical contribution in terms of theoretical construction is not clear, this paper has some merits in the system development and real application scenarios. In the survey of depth image acquisition, there are currently a number of new methods proposed. It is suggested to provide some newly information, such as the work "Depth Measurement Based on Stereo Vision with Integrated Camera Rotation," IEEE Transactions on Instrumentation & Measurement, Vol. 70, 2021. Since the image acquisition is an essential part of the system, it might provide more insight for the interested readers. 

Another fundamental issue of this work is to deal with the rendering from different viewpoints. There are many computer graphics techniques suitable for this requirement. Although the image based rendering is one of the mainstream methods, the authors also address the 3D based technique. In this paper, the deformation of human face seems not explicitly considered. This aspect should be discussed, for example, the method in “Novel view image synthesis based on photo-consistent 3d model deformation,” International Journal of Computational Science and Engineering, vol. 8, no. 4, 2013. 

Finally, the datasets used for evaluation is extremely important. Please make the description more clear, and have the datasets available if possible.

Author Response

Coimbra, 30th of August 2021

Dear Sir

The authors would like to thank the reviewers and their comprehensive comments that definitely contributed to achieve a much better version of the submitted article.

We addressed all the comments in the revision and submitted a version of the paper in track changes (using Microsoft Word tools) and the PDF paper where we highlighted the most important changes applied to the paper. We also gone through the paper with an extensive English revision.

In the highlighted PDF version of the paper, we highlighted in yellow new paragraphs and in blue moved paragraphs.

We hope that this version will be considered by the reviewers as publishable.

The detailed discussion of the reviewer comments can be read bellow.

REVIEWER 3:

“This paper presents a system to enhance the ID card validation using 3D visual information. An image is captured and several viewpoint rendering is generated for face recognition. The system implementation pipeline is described and experimental results are demonstrated. 

 Although the technical contribution in terms of theoretical construction is not clear, this paper has some merits in the system development and real application scenarios. “

We believe that the paper in the current version is much improved, and we expect it to have achieved a publishable version. The theoretical contribution is now more clearly stated. The current article intends to be a system article where the whole end-to-end system is presented and demonstrated.

“In the survey of depth image acquisition, there are currently a number of new methods proposed. It is suggested to provide some newly information, such as the work "Depth Measurement Based on Stereo Vision with Integrated Camera Rotation," IEEE Transactions on Instrumentation & Measurement, Vol. 70, 2021. Since the image acquisition is an essential part of the system, it might provide more insight for the interested readers.“

We provided more detailed explanations on the depth reconstruction including some more references as suggested. The PDF version of the paper highlights the most extensive changes and the Microsoft Word .docx file has all the changes made.

“Another fundamental issue of this work is to deal with the rendering from different viewpoints. There are many computer graphics techniques suitable for this requirement. Although the image based rendering is one of the mainstream methods, the authors also address the 3D based technique. In this paper, the deformation of human face seems not explicitly considered. This aspect should be discussed, for example, the method in “Novel view image synthesis based on photo-consistent 3d model deformation,” International Journal of Computational Science and Engineering, vol. 8, no. 4, 2013. “

This topic was better discussed, and the suggested article was cited, improving the reading easiness. The PDF version of the paper highlights the most extensive changes and the Microsoft Word .docx file has all the changes made.

“Finally, the datasets used for evaluation is extremely important. Please make the description more clear, and have the datasets available if possible.”

A paragraph on the datasets was added to the text. We discuss the datasets obtained and the inhouse constructed one. We will apply for the possibility to make the inhouse dataset publicly available after the paper publication.

End of revision comments

One more time, thanks to the reviewers for their work.

With our best regards,

Leandro Dihl

Leandro Cruz

Nuno Gonçalves.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

• There seem to be no significant advancements/improvements made in terms of the experiments carried out.
• The result section has no changes and seems to be a major concern. 
•  Section 3.1.1 still seems to be not necessary and it has no impact on the article. 
• What are the significant contributions presented in this article compared to that of the author's previous work? It is still not clear. 

Author Response

Coimbra, 10th of September 2021

Dear Sirs

The authors would like to thank the reviewers and their comprehensive comments on round 2 that contributed to achieve an even better version of the submitted article.

We addressed all the comments in the revision and submitted a version of the paper in track changes (using Microsoft Word tools) and the PDF paper where we highlighted the most important changes applied to the paper.

In the highlighted PDF version of the paper, we highlighted in yellow new paragraphs.

We hope that this version will be considered by the reviewers as publishable.

The detailed discussion of the reviewer comments can be read bellow.

REVIEWER 2:

“There seem to be no significant advancements/improvements made in terms of the experiments carried out.

The result section has no changes and seems to be a major concern.”

We added a section in the experiments related to the 3D reconstruction and mesh generation for the several types of cameras. This part of the experiments was not covered in the previous versions of the article. With these experiments we cover the whole process starting from the photo acquisition to the view generation and card printing. The new paragraphs are highlighted in yellow.

“Section 3.1.1 still seems to be not necessary and it has no impact on the article.”

We, respectfully, don’t agree with the reviewer. The technology of the camera influences the 3D model of head and face and the cameras used to it must be described and the mesh generation from the 3D reconstruction must be particularized for each type of cameras. We believe that with the adding of the section 5.1 in the experiments, the section 3.1.1 makes sense.

“What are the significant contributions presented in this article compared to that of the author's previous work? It is still not clear.”

As replied in the answer to reviewers, the current article is a system article where the whole end-to-end process is presented and discussed, while the previous work published by the authors is focused on the filtering of 3D models for smoothing. This was clarified in the paper with the following explanations: “In summary, the motivation of our initial study was the research and development of an end-to-end application to produce polycarbonate ID and travel documents with visualization effects using lenticular lenses. In the current article we present not only the filtering of 3D models as presented in the author’s previous work \cite{ref-Dihl2019}, but indeed the whole system, from the acquisition of images to the printing of the cards, by analyzing each of the pipeline phases and by discussing available technology and options taken. This article thus seeks to be an inspiration for other engineering articles in the industry of security printing and authentication of persons.” This paragraph is highlighted in yellow in the PDF paper.

End of revision comments

One more time, thanks to the reviewers for their work.

With our best regards,

Leandro Dihl

Leandro Cruz

Nuno Gonçalves.

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

• The response and changes made are noted.

Author Response

Coimbra, 13th of September 2021

Dear Sirs

The authors would like to thank the reviewers and their comments on Minor Revisions (round 3) that contributed to achieve an even better version of the submitted article.

We addressed all the comments in the revision and submitted a version of the paper in track changes (using Microsoft Word tools) and the PDF paper.

We hope that this version will be considered by the reviewers as publishable.

The detailed discussion of the reviewer comments can be read bellow.

REVIEWER 2:

“The response and changes made are noted.”

and

“(x) Moderate English changes required”

We made a revision of the English to the parts of the paper, essentially the new section 5.1 added in round 2 revision. The whole paper was then revised by an English-native reviewer.

End of revision comments

One more time, thanks to the reviewers for their work.

With our best regards,

Leandro Dihl

Leandro Cruz

Nuno Gonçalves.

Author Response File: Author Response.pdf