Multiple Object Tracking in Drone Aerial Videos by a Holistic Transformer and Multiple Feature Trajectory Matching Pattern

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The article proposes an effective new algorithm for recognizing groups of objects, which is necessary for orienting and achieving flight goals by a group of unmanned aerial vehicles. The article gives a sufficient introduction with an overview of the methods and an analysis of the advantages and disadvantages of known methods. Based on an analysis of sources and a study of known methods, including through their application, a new original method was proposed, developed and tested, the advantages of which were demonstrated using illustrative examples.

Unfortunately, when describing the results, the authors are too laconic; illustrations and tables are presented. Figures 11 and 12 are a series of illustrations using three methods, two pictures are given for each method, for a total of 12 illustrations. Apparently, the authors, trained in analyzing such illustrations, can easily see from the images presented the advantages and disadvantages of various methods, but it is not clear to the reader in what way exactly the pictures on the right differ from the pictures on the left and in the center. In such cases, it is customary in scientific articles to verbally describe the illustrations in terms of the differences between similar images, to indicate which features in similar, but in some way clearly different illustrations provide the basis for the conclusions that are made at the end of the article.

It is proposed to add appropriate comments to Figures 11 and 12.

This article does not require additional reviewing and after these minimal modifications it can be published.

Perhaps the name of one of the authors is misspelled, maybe instead of “ Yubin” YUan " should be written " Yubin Yuan "?

Author Response

Dear reviewer,
     Thank you for reviewing our paper and providing valuable feedback. We have revised your comments item by item, please refer to the attachment for specific details. Thank you very much for your guidance and suggestions, which have played an important role in improving the quality of our paper.
     If you have any further suggestions or questions, or if you require additional revisions, please don't hesitate to let us know. We greatly appreciate your assistance and support and look forward to hearing your further feedback.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The aim of this paper is the realization of an algorithm able to face the challenges like occlusion, scale variations and rapid motion that occur when the potential of drone surveillance is used to track multiple object. The authors propose a drone multiple object tracking algorithm based on Holistic transformer and multiple feature trajectory matching pattern. The trackers includes three components. By a comparative validations the authors are able to affirm their algorithm’s effectiveness and its potential into practical engineering applications.

Specific comments referring to line numbers, tables or figures that point out inaccuracies within the text or sentences that are unclear.

-          Please include a table with acronyms and in particular in the abstract write the acronym of GAO

-          Figure 1 and Figure 2 should have been called after its reference in the text. In this way it can be better understandable.

-          (line 172-174) if it is possible to represent in a scheme or in a figure what is explained in that sentence lines, could be better.

-          (line 205) equation 1 is never cited in the text. The same is for the equations from 2 to 7.

-          Please, add quantitative results in the abstract

-          Please, add the motivation of the work in the introduction

-          Please, add the organization of the paper at the end of the introduction

-          Please, add a brief introduction in the section 2 in order to highlight the object of the subsections

-          I suggest to move the contributions presented in the introduction, at the end of section 2 after the related work in order to assess the novelty compared to the existing literature

-          For each novelty (main contributions in section introduction) aimed by the authors, explicit explanation must be provided in the core of the paper

-          Please, include a discussion section

-          Please, extend the conclusion

 

Comments on the Quality of English Language

-          (line 22) Please, leaves a space between the reference and the last word of the sentence. It is like this throughout the text.

-          (line 22) “d” od drone in capital letter

-          Please take care of the space between the last word in a sentence and the reference in brackets. There is not always space.

-          line 436, space after “paper”.

-          line 440 a space between “YOLO” and previous word.

-          Line 443 “The”, please put a space before it.

-          Line 445 space before “VITDet”

-          Line 446 space after “RTD”

-          Line 449 space before “DN”

-          Line 472 please check the space before the capital letter

 

Author Response

Dear reviewer,
     Thank you for reviewing our paper and providing valuable feedback. We have revised your comments item by item, please refer to the attachment for specific details. Thank you very much for your guidance and suggestions, which have played an important role in improving the quality of our paper.
     If you have any further suggestions or questions, or if you require additional revisions, please don't hesitate to let us know. We greatly appreciate your assistance and support and look forward to hearing your further feedback.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Q1：

Section 3.1 of the article describes“In each stage, the Holistic Transformer block consists of 2, 2, 18, and 2 Holistic Transformer layers, respectively.” How is this number confirmed?

Q2：

What is the GT definition of MOTA in formula (25)? Please provide the full English abbreviation meanings in the text. Such as“Visual Gaussian mixture probability hypothesis density (VGM-PHD)”

Q3：

The article presents multiple algorithm comparison tables. Please elaborate on the concepts of the data indicators that are being compared across these tables throughout the article."

Q4：

 

Section 4.4.2 of the article describes“To ensure consistent comparisons despite variations in object distributions across datasets, we employed the Holistic Trans-Detector detector to produce uniform detection results for all tracking comparison methods”.The Holistic Trans-detector is an innovative method proposed in this paper, and the authors need to compare and validate this extraction method.

Comments on the Quality of English Language

 Minor editing of English language required

Author Response

Dear reviewer,
     Thank you for reviewing our paper and providing valuable feedback. We have revised your comments item by item, please refer to the attachment for specific details. Thank you very much for your guidance and suggestions, which have played an important role in improving the quality of our paper.
     If you have any further suggestions or questions, or if you require additional revisions, please don't hesitate to let us know. We greatly appreciate your assistance and support and look forward to hearing your further feedback.

 

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

It provides an excellent response to the posed question.

Comments on the Quality of English Language

Moderate editing of English language required