CGAN-Assisted Renovation of the Styles and Features of Street Facades—A Case Study of the Wuyi Area in Fujian, China

Round 1

Reviewer 1 Report

The research topic is very interesting and controversial at the same time. The issue of machine learning and artificial intelligence replacing architects will remain debatable. I think creativity will always be exclusive to humans. In addition, it should be clear in your paper that the role of artificial intelligence is to recommend options of facades; these tools are just a steppingstone for efficient designs. Thus, In your argument construction, this should be clarified

-        I think the authors should reconsider changing the title as it is more like a sentence than a title

-        the manuscript has many English language and editing issues, as it has many spelling and grammar mistakes. for instance, “data capturing”not “capturing” in line 137, and it’s advisable to use passive voice in scientific writings. ( he learned line 140, we redrawed line150,….)

Concerning the abstract line 23,24 …” (2) In terms of product graphic design, espe-  cially the common tourist cultural products fans and water cups, show significant potential.”

I think graphic design is a bit irrelevant to the scope of the research.

 

-        Concerning Figure 4, The experimental materials, it should be written clearly if this part is concerned with Model training. In addition, How the CGAN works gets clearer in the text following this section. I think this should appear earlier so that the reader doesn’t get confused.

 

-        In line 256, Under the title of Model application, you wrote that three main research problems, one of them was “to increase the facade design according to the opening of doors and windows for buildings lacking facade decoration.”, this part needs clarifying as the opening in architecture has a major role concerning the indoor- outdoor relationship, daylighting and ventilaton and environmental factors so it’s not advisable to increase them with no good reason, just for decorating the façade. If you mean adjusting or modifying the openings style, this should be clarified

 

-        Concerning the part of the application of the model on village buildings, there are no significant changes in the facades done by CGAN. I think these slight changes could be done by architects easily.

 

-        While the part of unfinished buildings in the village, how could the CGAN suggest the location of the opening while there are no parameters determining the spaces’ function lying behind these external walls, there are no considerations for the windows size and orientation and its relation with privacy, ventilation and daylighting

Author Response

Comments:The research topic is very interesting and controversial at the same time. The issue of machine learning and artificial intelligence replacing architects will remain debatable. I think creativity will always be exclusive to humans. In addition, it should be clear in your paper that the role of artificial intelligence is to recommend options of facades; these tools are just a steppingstone for efficient designs. Thus, In your argument construction, this should be clarified

- I think the authors should reconsider changing the title as it is more like a sentence than a title

Response: Yes. I agree with your statement. Traditional computers or artificial intelligence software cannot completely replace architects. Our previous headline really didn't express it well. The current title of the paper has been revised. Research on Machine Learning-assisted the Architectural Facade Styles Renovation Design—A case study of Wuyi area in Fujian, China

 

Comments:- the manuscript has many English language and editing issues, as it has many spelling and grammar mistakes. for instance, “data capturing”not “capturing” in line 137, and it’s advisable to use passive voice in scientific writings. ( he learned line 140, we redrawed line150,….)

Response: We are also very sorry! Cantonese is our primary language. In terms of English expression, it has already been handed over to the English proofreading and polishing service provided by MDPI for further processing. We take this issue very seriously! Thank you very much for pointing it out too!

 

Comments:Concerning the abstract line 23,24 …” (2) In terms of product graphic design, espe-  cially the common tourist cultural products fans and water cups, show significant potential.”

Response: Yes, the sentence of the abstract corresponds exactly to the part of the article in addition to the auxiliary renovation design of the building facade, as well as the derivative tourism products. At the same time, we also made revisions to the Abstract section.Of course, the language proofreading part is further improved by using MDPI's author service. Thank you for reminding!

 

Comments:I think graphic design is a bit irrelevant to the scope of the research.

Response: Dear professor, we can tell you very honestly. The Wuyi architectural style is a unique regional architecture in the Wuyi Mountain area. The design of the product is just to illustrate an additional discovery that our previous building facade design meets. Just to show that this method has a wider range of applicability. What they have in common is that they all belong to the derivatives of Wuyi architectural style. At the same time, we have made further improvements to the relevant text in the text.

 

Comments:- Concerning Figure 4, The experimental materials, it should be written clearly if this part is concerned with Model training. In addition, How the CGAN works gets clearer in the text following this section. I think this should appear earlier so that the reader doesn’t get confused.

Response: Your suggestion is very good. We have made further improvements in the text.

 

Comments:- In line 256, Under the title of Model application, you wrote that three main research problems, one of them was “to increase the facade design according to the opening of doors and windows for buildings lacking facade decoration.”, this part needs clarifying as the opening in architecture has a major role concerning the indoor- outdoor relationship, daylighting and ventilaton and environmental factors so it’s not advisable to increase them with no good reason, just for decorating the façade. If you mean adjusting or modifying the openings style, this should be clarified

Response: We modified it: In the case of this study, the project has three main problems to be solved: (1) How to preserve the regional style in the design of the new building. (2) How to modify the ex-terior decoration of unadorned structures according to the existing door and window openings. (3) How to restore unfinished buildings according to regional style. In addition, we propose that the model can be applied to some cultural products in villages and towns, such as road signs, house numbers, and souvenirs.

 

Comments:- Concerning the part of the application of the model on village buildings, there are no significant changes in the facades done by CGAN. I think these slight changes could be done by architects easily.

Response: Referring to the results shown in Figures 9 to 11, there are still differences in the building facades. Because we have divided different building types in it. Although the overall style of the building facade is the same, but due to the different types, the output results of the application part are also different.

 

Comments:- While the part of unfinished buildings in the village, how could the CGAN suggest the location of the opening while there are no parameters determining the spaces’ function lying behind these external walls, there are no considerations for the windows size and orientation and its relation with privacy, ventilation and daylighting

Response: In the village buildings in Wuyi area, due to the different nature of use, they can be divided into functional components and decorative components. In the overall renovation process of the village, in order to maintain the overall style, there are often some decorative components. That is to say, it is dominated by decorativeness. We have already considered the indoor layout and function of the sample materials when conducting machine learning training. After the training of machine learning is completed, the effect presented has certain reference and guidance for opening. Of course, this belongs to the presentation stage of conceptual deliberation.

Reviewer 2 Report

The topic of the paper is very relevant and in fact the application of AI technologies in architecture generation and especially in the heritage context currently are state-of-the-art research trends. However several aspects of research should be improved before publication:

- more in-depth literature review of existing similar studies of AI generated architecture especially in the heritage contexts must be presented at the same time highlighting the scientific novelty and relevance of this research in existing research context;

- the link with sustainability of this study should be demonstrated;

- the aim of research should be more clearly presented in the introductory section;

-  critical discussion of research results in the light of cultural expression, local identity, sprit of place must be presented, probably the SWOT analysis of the approach could be carried out.

Author Response

Comments:The topic of the paper is very relevant and in fact the application of AI technologies in architecture generation and especially in the heritage context currently are state-of-the-art research trends. However several aspects of research should be improved before publication:

- more in-depth literature review of existing similar studies of AI generated architecture especially in the heritage contexts must be presented at the same time highlighting the scientific novelty and relevance of this research in existing research context;

Response: Your suggestion is very good! Regarding the literature on artificial intelligence, we have searched and added again. "1.2 Literature review" expands on writing. Please refer to the revised section for details.

 

Comments:- the link with sustainability of this study should be demonstrated;

Response: Since China's Wuyi Mountain is a world cultural heritage and natural heritage, according to the current construction of national parks and national archaeological parks in China, there is a strong demand for the unity of the surrounding street facades. The state has continuous financial support for Mount Wuyi, mainly for the construction of architectural style control. The purpose of this study is to better improve the regional architectural style of Wuyi style, so that the protection of traditional buildings in the Wuyishan area can be strongly supported, and the architectural style with more regional characteristics can be completed, so as to enrich the material wealth of cultural tourism in the Wuyishan area. From the perspective of sustainable heritage protection and sustainable tourism development, this research is very relevant to sustainability. In the corresponding text part, we have carried on the perfection.

 

Comments:- the aim of research should be more clearly presented in the introductory section;

- critical discussion of research results in the light of cultural expression, local identity, sprit of place must be presented, probably the SWOT analysis of the approach could be carried out.

Response: Thanks a lot for your suggestion. We have made further improvements in the article. At the same time, please rest assured that we have chosen the author service of MDPI Publishing to proofread and polish the English expression. Try to achieve better expression.

Reviewer 3 Report

The paper is very interesting and very well structured. The conclusions are clear and the references adequate. I believe that the leterature review section must be enriched. The section "The above methods have shown  great potential in the field of architecture, such as building facade generation [11],  building volume generation [12], building floor plan optimization [13], street view quality assessment [14], street view image generation [15], Architectural style migration [16].  All of them can greatly improve the efficiency of building evaluation and building design.  Among them, in the field of building facade generation, the current research is  mainly based on the Generative Adversarial Network (GAN) [17]" must descibe in a clearer way the other methods mentioned and justify the selection of the GAN method.

 

Author Response

Comments:The paper is very interesting and very well structured. The conclusions are clear and the references adequate. I believe that the leterature review section must be enriched. The section "The above methods have shown  great potential in the field of architecture, such as building facade generation [11],  building volume generation [12], building floor plan optimization [13], street view quality assessment [14], street view image generation [15], Architectural style migration [16].  All of them can greatly improve the efficiency of building evaluation and building design.  Among them, in the field of building facade generation, the current research is  mainly based on the Generative Adversarial Network (GAN) [17]" must descibe in a clearer way the other methods mentioned and justify the selection of the GAN method.

Response: Your suggestion is very good! Regarding the literature on artificial intelligence, we have searched and added again. "1.2 Literature review" expands on writing. Please refer to the revised section for details.The corresponding serial numbers increase from [20]. The corresponding references have also been improved.

Round 2

Reviewer 1 Report

No further comments.