Indoor Distance Measurement System COPS (COVID-19 Prevention System)
Round 1
Reviewer 1 Report
The article presents Indoor distance measurement system for prevention of COVID-19. The structure of the paper is good. But to be publishable it needs major revision in presentation and to be more clear and usable for the researchers in the field. There are some issues that should be resolved to make the article acceptable, otherwise, the work should be improved and resubmitted as another submission. To improve the quality of the article, the authors should follow the given comments:# The abstract should be more detailed on the contributions of the presented work and should represent some specific result at the end. At the current form, the contribution of the work sounds very limited. # The presented work requires some numerical evaluation on the results as well, the article is more focused on how to carry on the work rather than giving some concrete numerical results. # In the introduction, one part should be added about research works carried out on social distancing. # A point of the work is using a CNN-based recognition system. Currently, deep learning-based recognition is gaining increasing attention, and it has been deployed in different fields, the authors should point to the importance of CNN in recognition applications by referring to applications like deep face recognition as in the research work "Deep Face Recognizer Privacy Attack: Model Inversion Initialization by a Deep Generative Adversarial Data Space Discriminator". # The section Related Works does not seem to discuss just the related works. the authors should review the related research works carried on the same target of the article. # As the application of smartphone has a key role in this research, the authors should mention the importance of the smartphone in IoT research in other fields as well, like "Hydraulic system onboard monitoring and fault diagnostic in the agricultural machine", "Economic data analytic AI technique on IoT edge devices for health monitoring of agriculture machines". # Regarding the pre-processing of the images, the authors can point to the strong nonlinear techniques for removal of possible noise from the images like Mediated Morphological Filters (2001). # The English of the article needs checking and improvement as in case of the sentence "The process of pairing, which must be performed to attain a connection between two devices, disappears, increasing the communication distance between devices, and the signal strength can be transmitted to enable accurate positioning" is very long, it should be divided to several sentences. Also, it is not clear what mainly trying to say.
Author Response
Thank you for your advice
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The article is unclear and inconsistent. There is no logical scenario of the application and it does not solve a real problem.
There are several premises that I consider wrong :
- How do you know a person is infected?
- How can an infected person be detected?
- Why does an infected person move freely?
- Why is it necessary to follow the infected person in public instead of stopping his public circulation?
- One assumes that an infected person automatically infects all the people against whom he/she is less than 2m
- Where are the outdoor CCTVs placed and how many webcams are needed in an open environment? Is it necessary to run the application in an outdoor environment?
- One assumes that each person has a smartphone.
The article has no theoretical contributions, but only describes at a conceptual level a system for tracking objects in the neighborhood of a reference object (a person assumed infected), by detecting them and measuring the distance between them and the reference object.
Basic information is unnecessarily detailed. The description of some basic operations and components is unnecessarily repeated (example fig. 6).
Other remarks:
Line 168 “ the figure on the left..” I do not understand what this figure is.
Line 174 “ the upper right picture..” is it not the original one?
Author Response
Thank you for your advice
Author Response File: Author Response.pdf
Reviewer 3 Report
The autors present an interesting paper about an indor distance measurement system, based-on YOLO, which the authors propose its application to motigate the spread of COVID, through the detection of thresholds associated with the distance between people. While the first part (distance estimation) is clearly described, as for the second part of problem (prevention of disease propagation, identification of "people", and subsequent tasks to be taken) it seems to merit a better theoretical fundation, both at conceptual level and in terms of the effective pratical appplication of this solution. Thus, authors should review the paper before publication. In addition, authors should also consider improving the following issues:
- pag4:Ln146-7: yes, the system seems to work as a tool to detect the contact within some predefined limits. But assuming that among these contacts there is a contaminated person, how are the other people identified, to be notified about it?
- it is not visible in the text any comments in way to describes the action/alert/notification between the intervenient people, or even some action of prevention/alert as measures that the distances between actors are decreasing ir order to reach the defined threshold value.
- pag5: the text of the description of the different components in Figure 3 do not match with the image
- pag13: Ln317-8: Is that so? What is the process of identifying effectively people so that thaey can later be contacted?
- Even if the authors intend to treat some of theses issues as future work, they should give some indication (more specific) on that subject in the conclutions sections.
Author Response
Thank you for your advice
Author Response File: Author Response.pdf
Reviewer 4 Report
Authors of this paper present an object detection based technique, employing video data collected by CCTV, for indoor tracking of persons. Wi-Fi fingerprinting is used to calibrate the measured distances. The proposed technique is motivated by reducing the coronavirus disease (COVID-19) that which can spread an infected person in an indoor environment. In my opinion, the content of the article does not meet well with the topics of the MDPI Sustainability journal. Journals like Sensors and Remote Sensing (both MDPI) should be better.
In general, the article is prepared on general level and many requirements on a journal paper are fulfilled. However, there are several some shortcomings, which must be fixed. For more details, please, see my comments!
Comments:
- Article – the English grammar contains some minor typos, for instance: line 42-46 “pollution. [7-9].”; line 171 “and d the”; line 263 “The values used are”; line 33 “2m”-it should be “2 m”; line 309 “traced I can”; etc. Please, check the whole article carefully once again! (In general, the article has a good readability.)
- Introduction – paragraph about the organization of the article is missing
- Introduction and Section 2 – the state-of-the-art (SOTA) is elaborated on general level in Section 2 and the main contributions of the article are defined in Section 1. I recommend for the authors to more highlight the main contributions of the article related to SOTA. The using of items is recommended.
- Section 2 – there are briefly discussed some techniques developed for indoor localization and tracking. There also parts focused on BLE and Wi-Fi RSSI-based localization. Some links on appropriate works should be nice to present here, e.g. “Do we need a contact tracing app?”; “BLE device indoor localization based on RSS fingerprinting mapped by propagation modes” and “Adaptive distance estimation based on RSSI in 802.15.4 network”. The first one presents an interesting study (contact tracing by smartphone for the containment of the COVID-19 pandemic). The second one evaluates the BLE localization performances based on RSS fingerprinting (based on extensive indoor measurements), mapped by propagation modes, in indoor environment. Finally, the third one deals with the RSSI distance estimation issue in wireless sensor networks. I hope that the mentioned papers can be helpful for you. Otherwise, please find other ones!
- Section 2 – what is the main difference (or differences) between the proposed system structure (Fig. 1) and previous SOTA solutions?
- Section 4.1 – it is written: “Wi-Fi has a frequency band, which we set to perform Wi-Fi functions.” Please, explain it in details!
- Section 4.3 – Fig. 5 has very bad visibility (bad sizes). Please, fix it! Next, this figure is not discussed (described) in details. Please, check it!
- Article – sometimes, the figure is presented before its mention (“introducing”) in the text. Firstly, the figure should be mentioned in the text and after that presented. If it is possible, please fix “typo”! Thanks!
- Section 4.5 – please, check the sentence at lines 213-215!
- Section 4.6 – equations (1) and (2) in the pdf document look like a blurred pictures! Please, check and fix this typo!
- Section 4.5 – please, give more details about the utilizing of the CNN-LSTM model! How many percent of data are used for training the model?
- Section 4.7 – it is written “The distances between objects measured through 4.6 are identified” – ??
- Section 4 – for reproducible research, I would suggest the authors make the code of their technique (signal processing, e.h. for CNN-LST) and dataset publicly available if possible.
- Section – the performance of the proposed technique is evaluated on general level. Its comparison with similar SOTA solutions, if possible, should be presented in an appropriate section.
- Section 5 – how do you plan the test of your proposed solution in real conditions?
Author Response
Thank you for your advice
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The revisions in the article have not been reflected in the response. The authors' acts in response to each comment should be mentioned in the response to the reviewer, and just saying the "We have revised based on thecomment" is not enough.
In addition, as reviwer went through the revised article, some of the comments of the Reviewer have not been followed as non of the works mentioned in the comments can be found in the article. It is suggested the authors apply the reviewers comments from the first round of the review carefully to make the work closer to publication.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 2 Report
The paper was improved with respect to the first manuscript.
Please, explain how you can distinguish between an infected and a non-infected person.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 4 Report
The article has been partly improved. Many thanks for the explanation letter! However, many times, the response of authors on the comments is confusing and does not complete (one or two sentences – what I the meaning of “We have deleted what you said as the content changed”?. The explanation letter does not contain links on the article – where the changes are visible. Next time, please prepare your explanation letter better! Thanks!
After the check of the article, I have the following comments:
- Article – the article still contains some typos in the English grammar (e.g. “the devices. [26].”; it should be “CNN (Convolution Neural Network)”; “I think that with tools and…” – it is very unoformal; etc.)
- Introduction – I still recommend for the authors to more highlight the main purposes/contributions of the article in more visible form – e.g. by employing of items
- Table 1 – I recommend for authors to use words “Yes” and “No” instead of “circle” and “X”, respectively. Next, the order of “Paper Authors” is wrong – e.g. “Cupta, S. [33]”
- Article – please, extend your article with your response on my comment n. 5 (see the explanation letter)
- Table 3 –please, improve the presentation of Table 3, mainly the column “List of coordinates” (brackets should not be divided)
- Article – I have checked the proposed concept once again. I have this question: How do you know which person is infected?
- Section 4 – for reproducible research, I would suggest the authors make the code of their technique (signal processing, e.g. for CNN-LST) and dataset publicly available if possible
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 3
Reviewer 1 Report
As the authors have revised the paper well and followed the reviewers' comments in detail, the work is qualified for publication.
Reviewer 4 Report
The article has been improved. Thanks for the explanation letter!
