Global-Local-Structure Collaborative Approach for Cross-Domain Reference-Based Image Super-Resolution

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript presents a novel framework for reference-based remote sensing image super-resolution, integrating degradation-aware modeling, a dual-decoder recursive generation strategy, and static regularization guidance. The approach is motivated by clear limitations in existing methods, and the experimental results demonstrate improved performance under both ideal and blind degradation scenarios. The work is relevant and contributes to the field of remote sensing image enhancement. However, the manuscript requires revisions in several areas to enhance clarity, methodological explanation, and overall presentation before it can be considered for publication.

Major Comments:  

1. The manuscript would benefit from thorough language editing to improve sentence fluency and logical flow, particularly in the Introduction and Conclusion sections. Some paragraphs are repetitive (e.g., multiple discussions on degradation modeling), and consolidating these would strengthen the narrative.
2. Several figures referenced in the text (e.g., Fig. 6, Fig. 7) are not included in the submitted draft. These must be provided to support the qualitative analysis.
3. There are inconsistencies in formula numbering and citation (e.g., Equation (4) is referenced but not presented). Please verify all equation references align with the numbered equations in the text.
4. While quantitative results are provided, the discussion of perceptual metrics (LPIPS, VIF) is limited. A brief explanation of their relevance to remote sensing image quality assessment would be helpful.
5. Computational efficiency (e.g., inference time, model parameters) is not discussed. Including such details would better highlight the practical applicability of the proposed framework.
6. Reference formatting is inconsistent (e.g., journal abbreviations, conference names). Please ensure uniformity according to the journal’s style guide.
7. Several recent works (2024–2025) are cited, but a more thorough comparison with contemporary diffusion-based SR methods in remote sensing would better contextualize the contributions.

Minor Comments:

1. Some acronyms are introduced without full definitions (e.g., SRG, LGDF). Ensure all are clearly defined upon first use.
2. The captions for Tables 1 and 2 should explicitly state that higher values are better for PSNR, SSIM, VIF and lower for LPIPS, RMSE.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript proposes a novel hierarchical multi-task super-resolution framework to match local detail preservation as well as global structural consistency for image super-Resolution. The work is innovative and experiment results on two datasets illustrate that the proposed method achieves superior performance. However, there are several problems that need further consideration and improvement by the author, as follows:

 

1. The motivations and contributions of the manuscript are not sufficiently described in the introduction.
2. Remote imagery has a wide range of applications. While the current focus on super-resolution is valuable, discussing other applications in the Introduction would help broaden the scope and impact of the study, such as detection and classification tasks, etc.
3. Equation (28) defines the loss function composed of three terms, with the corresponding weights set to 1.0, 0.1, and 1.0. However, the manuscript does not clearly explain the rationale for setting these weights. I recommend conducting a parameter sensitivity analysis to show how different weight settings affect the performance.
4. It is well-known that the remote sensing image tends to suffer from various degradation, noise effects, or variabilities in the process of imaging. The reviewer is wondering what will happen if the proposed method meets the various variabilities.
5. To provide a more comprehensive comparison, it is suggested to report and analyze the inference time and model parameters of the proposed method against other state-of-the-art methods.
6. Line 8 of the manuscript contain a spelling error (“dual-doder platform”) and should be corrected.
7. The English grammar and expression could be improved with the help of professionals. There are many long sentences which are not easy for readers to understand.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

In the manuscript, a novel hierarchical multi-task super-resolution network is proposed for image super-resolution. Experiments show the effectiveness of the proposed method. However, there are some problems requiring further solving, as follows:

1. The related work of the manuscript would benefit from incorporating additional recent studies, particularly those published in the last two years.
2. At the end of the introduction, please add a brief summary of the paper’s overall structure. This would guide the reader and make the flow of the manuscript clearer.
3. More SOTA methods published in recent two years should be used as the comparison methods.
4. It is recommended to include the model parameter count and FLOPs for completeness.
5. The experiment section describes a lot of accuracy comparison results and descriptions, but the reason why the proposed innovations make the accuracy increase is inadequate.
6. The manuscript does not analyze the potential limitations of the proposed method.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

1. The overall framework of the paper is well structured; however, the core innovations and distinctions of the proposed method with respect to existing degradation-aware diffusion-based super-resolution approaches should be more clearly articulated.
2. The design motivation and working mechanism of the degradation-aware modeling module (DAM) are not sufficiently explained, and the conceptual explanation in the methodology section should be further strengthened.
3. The relationship between the claimed “single-step diffusion” and the recursive generation mechanism is described in a somewhat ambiguous manner, and the actual inference process and computational characteristics should be clarified.
4. While the experimental comparisons are generally comprehensive, some baseline methods are not fully aligned in terms of task setting and prior assumptions, and the fairness of the comparisons should be more clearly discussed.
5. Although the ablation studies demonstrate the effectiveness of individual components, the discussion on key design choices could be further improved to strengthen the overall argumentation.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This paper has addressed all my concerns. I think the current version of the manuscript can be considered for publication.

Author Response

We sincerely thank the reviewer for their careful evaluation and constructive feedback. We are grateful for the positive assessment and for recognizing that our revised manuscript addresses all concerns.

Reviewer 3 Report

Comments and Suggestions for Authors

I have no more comments.

Author Response

We appreciate the reviewer’s time and thorough review of our work. We are pleased that the revised manuscript meets your expectations and resolves all comments.