Single-View 3D Reconstruction via Differentiable Rendering and Inverse Procedural Modeling^†

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors,

1. The Method overview is not so clear to be replied.

2. Please, present the process workflow.

3. Eq. 4 to Eq. 8 need to be better explained and their applications. They are not so clear.

4. Why and how did you define the hardware?

5. Present the comparison between algorithms in Table 1 previously.

6. At the end of the section 4, you are advised to show a brief dicussion instead in section 5.

7. The figures need to present the differences more clearily.

Author Response

The answer to reviewer is in attached pdf.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors proposed a model to enhance the performance of 3D model reconstruction. There are some key comments to be considered by the authors before the recommendation of acceptance. A major revision is thus recommended. Please refer to my comments below.

Comment 1. Affiliations: Include information related to the School/Faculty/Department.
Comment 2. The paper title does not precisely describe the key techniques, e.g., “… differentiable rendering and inverse procedural modeling”, and “differentiable and non-differentiable procedural generators into a single framework that allows us to apply inverse procedural modeling to fairly complex generators”.
Comment 3. Abstract:
(a) Were two options adopted?
(b) Briefly describe the research results and implications.
Comment 4. Keywords: More terms should be included to reflect the scope of the paper better.
Comment 5. Section 1 Introduction:
(a) Elaborate on the importance of the research topic.
(b) Regarding three research contributions:
- Elaborate on the novelty and benefits of the approach in the first point.
- Second point, what is “to solve this problem”?
- Last point, explicitly share numerical results.
(c) Elaborate on the novelty and contributions compared with conference papers [8,9].
Comment 6. Section 2 Related works:
(a) Add an introductory paragraph before Subsection 2.1.
(b) Ensure all cited works were related to 3D model reconstruction.
(c) Ensure methodology, results, and limitations of cited works are described.
Comment 7. Section 3 Method overview:
(a) The heading of the section is not appropriate. If it is an overview, only high-level architectures/diagrams and their details should be presented.
(b) Incomplete discussion on each part of the technique/algorithm. Ensure equations/pseudo-codes/algorithms are presented to facilitate the rework of the implementation.
Comment 8. Section 4 Results:
(a) Add an introductory paragraph before Subsection 4.1.
(b) Besides figures, more numerical results should be presented.
(c) Ensure in-text citations are presented for those existing works being compared.
(d) In Section 2, the authors cited many related works. However, most of the works are not compared in Section 4.

Comments on the Quality of English Language

Minor spell check and typos. In addition, ensure proper spacing between wordings/symbols. 

Author Response

The answer to reviewer is in attached pdf.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have significantly enhanced the paper’s quality. I have some follow-up comments.
Follow-up comment 1. The first paragraph of the introduction: Justify the advantages of the procedural approach.
Follow-up comment 2. “We further describe of what has already been done before us in these areas and finalize motivation for our research in 2.7.”, please write “Subsection 2.7”.
Follow-up comment 3. Table 2, existing work [30] outperformed the authors’ work. What is the contribution of authors’ work?
Follow-up comment 4. Conclusion: Discuss the research limitations of the authors’ work. Elaborate on future research directions.

Author Response

Author's response is in attached pdf.

Author Response File: Author Response.pdf