Generating Structurally Complete Stylish Chinese Font Based on Semi-Supervised Model
Round 1
Reviewer 1 Report
The authors consider the problem of generating a structurally complete stylish Chinese fonts. To address the structural incompleteness problem, they propose a semi-supervised model that incorporates stroke encoding and an additional attention module. The proposed method has shown its effectiveness in comparison with known methods.
The article is well structured and written in a good language.
Some notes:
1) Extra parenthesis in the keywords.
2) Authors claim that “This paper presents a novel semi-supervised approach for generating stylish Chinese fonts”. However, some relevant related works are not discussed in the article, for example: https://arxiv.org/abs/2211.06198
3) Please, inform the reader about the structure of the paper at the end of Introduction section
4) Line 228 “…the weight relationship between the two, we limit the µ coefficients…” – a word missing
5) Equation (5) – The bottom dot sign is unclear here
6) Line 286 - what is meant by “number size”?
7) It will be better to present Table 1 as a figure
8) Line 326 - “Then the effectiveness of the improved attention module and the combination of the semi-supervised scheme and stroke encoding through ablation experiments.” – Unclear sentence
9) Line 339 – Figure 7 should be instead of Figure 6.
10) All figures and tables should be mentioned in the text before the figure (table) appears.
In general, the article has a good level of English. A few minor issues are shown below:
Line 228 “…the weight relationship between the two, we limit the µ coefficients…” – a word missing
Line 286 - what is meant by “number size”?
Line 326 - “Then the effectiveness of the improved attention module and the combination of the semi-supervised scheme and stroke encoding through ablation experiments.” – Unclear sentence
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
1- It is preferable that the summary and conclusions contain numerical values for the metrics that demonstrated the efficiency of the system
2- No reference was made to other techniques in the related work section, such as machine learning, to demonstrate the time progression of using deep learning techniques.
3- It is preferable to add an applied example to illustrate the work of the proposed algorithm
4- Standardize the format of references. In addition to adding two recent references published during the year 2023.
5- Is there any recent previous work in the same field to compare with?
6-Is it possible to add clarification about the time it takes to implement the system during training or examination?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf