Real-Time Object Detection in Remote Sensing Images Based on Visual Perception and Memory Reasoning

Round 1

Reviewer 1 Report

Good paper.

Section "Experiment" should be divided: "Results" and "Discussion"

Minors:
Remove background gradient (Fig.1, Fig.2, Fig.3, Fig.5, Fig.7)
Improve contrast in Fig.7

Fig.2 text in image too small
Fig.4 black text color required
Fig.4c text in image too small
Fig.5 too small

Eq.1 Eq.2 split to two subformulas, e.g. (1a) and (1b)

Eq.12 Eq.13 numerator and denominator format should be used

Eq.18 '<=' operator should be corrected for two first lines

l.220 Reference <- reference

l.353 brackets to fix

There are a lot formula formating problems.

 

 

Author Response

Author's Notes to Reviewer

 

Dear Reviewers:

Thank you for your comments concerning our manuscript entitled “Real-time Object Detection in Remote Sensing Images Based on Visual Perception and Memory Reasoning” (ID: electronics-614457). Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval. The main corrections in the paper and the responds to the reviewer’s comments are as flowing:

 

Reviewer’s comments:

Good paper.

 

Section "Experiment" should be divided: "Results" and "Discussion"

 

Minors:

Remove background gradient (Fig.1, Fig.2, Fig.3, Fig.5, Fig.7)

Improve contrast in Fig.7

 

Fig.2 text in image too small

Fig.4 black text color required

Fig.4c text in image too small

Fig.5 too small

 

Eq.1 Eq.2 split to two sub formulas, e.g. (1a) and (1b)

 

Eq.12 Eq.13 numerator and denominator format should be used

 

Eq.18 '<=' operator should be corrected for two first lines

 

l.220 Reference <- reference

 

l.353 brackets to fix

 

There are a lot of formula formatting problems.

 

 

 

 

 

 

 

Responds to the reviewer’s comments:

 

Thank you very much for your review of this paper. Your opinion is very valuable to us.

According to your opinion, we have divided the Section "Experiment" into two parts: "Results" and “Discussion".

We have removed background gradient (Fig.1, Fig.2, Fig.3, Fig.5, Fig.7) and Improved contrast in Fig.7.

We have greatly improved the quality of Fig.2, Fig.4, Fig.5 to make it easier to understand.

We have also revised the format of the formula you proposed, we hope you will be satisfied with it.

This paper has been polished by professional organizations to correct wrong expressions, typos or other errors.

We tried our best to improve the manuscript and made some changes in the manuscript. These changes will not influence the content and framework of the paper. We appreciate for your warm work earnestly, and hope that the correction will meet with approval. Once again, thank you very much for your comments and suggestions.

Author Response File: Author Response.pdf

Reviewer 2 Report

The article describes a new architecture of deep learn neural network for object detection in high resolution air surveillance pictures. Advantages of the article: 1) A new architecture od DLNN is proposed. 2) The architecture is implemented on TensorFlow library. 3) Tests was conducted on publicly obtainable data sets. 4) The architecture has superior properties to other architectures and algorithms known from literature. 5) Excellent literature review is given. Disadvantages of the article: 1) Reader have a little chance to reproduce computations conducted by authors.

Author Response

Author's Notes to Reviewer

 

Dear Reviewers:

Thank you for your comments concerning our manuscript entitled “Real-time Object Detection in Remote Sensing Images Based on Visual Perception and Memory Reasoning” (ID: electronics-614457). Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval. The main corrections in the paper and the responds to the reviewer’s comments are as flowing:

 

Reviewer’s comments:

 

The article describes a new architecture of deep learn neural network for object detection in high resolution air surveillance pictures. Advantages of the article: 1) A new architecture od DLNN is proposed. 2) The architecture is implemented on TensorFlow library. 3) Tests was conducted on publicly obtainable data sets. 4) The architecture has superior properties to other architectures and algorithms known from literature. 5) Excellent literature review is given. Disadvantages of the article: 1) Reader have a little chance to reproduce computations conducted by authors.

 

 

Responds to the reviewer’s comments:

 

Thank you very much for your review of this paper. Your opinion is very valuable to us.

Thank you for your recognition of our work. Our algorithm uses open data sets and open source libraries. Because we are currently applying for a software patent for our algorithm, we are temporarily unable to disclose our code.  After we have completed the patent application, we will open up our code on blogs and GitHub to facilitate colleagues to reproduce the algorithm.

This paper has been polished by professional organizations to correct wrong expressions, typos or other errors. We tried our best to improve the manuscript and made some changes in the manuscript. These changes will not influence the content and framework of the paper. We appreciate for your warm work earnestly, and hope that the correction will meet with approval. Once again, thank you very much for your comments and suggestions.

Author Response File: Author Response.pdf