Review Reports

Round 1

Reviewer 1 Report

I appreciate the authors' contribution to the novel formulation and algorithm. However, I don't understand the motivation of this research because I don't see the meaning of ``groups of sparse representation." Could the authors clarify the meaning of ``groups" in this context? Also, could the authors explain how the grouping improves the accuracy in the UAV123 data analysis?

I can't see the superiority of the authors' proposed method over the existing techniques from Fig.3 - 5. The authors should describe some considerations for the results from Fig.3 - 5. 

I found several spelling mistakes throughout the paper. It is better to use grammar check software.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript proposed a new design of sensing matrix and applied it to the UAV images. The proposed model achieves good performance in a series of experiments. The paper is well-organized and written clearly. Despite a rather clean presentation, some minor comments should be addressed.

1. It is more proper to consider that the paper proposed an image recovery framework based on compressive sensing and used some UAV image datasets in numerical experiments. Other than that, I don’t see a close relationship between the proposed algorithm and its application to UAV images. For example, an online learning framework is more related to UAV images. 

2. Since f is a least-squares problem w.r.t. Phi^T Phi in (19), the related lengthy derivations could be much more concise.

Some typos:

1. the authors seem to use \xi and \eta interchangeably  

2. "the critical issue of our research." -> "the main focus of our research"

3. "by employed the CS technique" -> "by employing the CS technique"

4. "low dimensional measurement" -> "low dimensional measurements"

5.  "constraint of sparse" -> "constraint of sparseness"

6. Constraint on Gt is missing, resulting to trivial solutions: Gt=0, Phi=0

7. "focused on how does" -> "focused on how the ..."

8. G in (7) is not defined

9. "controlled being projected" -> "controlled when being projected"

10. "more accuracy" -> "high accuracy"

11. =min ->  = arg min

12. authors proposes -> authors propose

13. "which extracting 15 patches" -> "which extract 15 patches"

14. (33)  || ||_2^2 should be || ||_F^2

15. with different dimension -> with different dimensions

16. are also list -> are also listed

17. are match with -> are matched with

Author Response

Please see the attachment.

Author Response File: Author Response.pdf