TwSense: Highly Robust Through-the-Wall Human Detection Method Based on COTS Wi-Fi Device
Round 1
Reviewer 1 Report
-The related work is not up-to-date since the latest reference is merely up to 2021. -The related work is not updated since the last reference is only until 2021. There are new works from the year 2022.
-The authors should clearly state the innovation and novelty of their work, while the significance of the results should be discussed more.
-Introduction may highlight the problem definition of the paper and main contributions of the paper.
-The description of the algorithm is too simple. The key variables, theories and process are not depicted in a professional way, which impact the accurate understanding of the read. All the mathematical equation used in this paper need proper citation with clear meaning of the variable used. All the mathematical equation used in this paper need proper citation with clear meaning of the variable used.
-The English of this paper should be polished and revised carefully
Author Response
Please refer to the attached document.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
1. The fields of equations are not mentioned, and the work accompanies the symbol table for better accessibility of symbols.
2. there is nothing in Fig. 3a
3. please explain Fig. 13 in a better way, you have not cleared why are you giving this comparison in your paper. the given explanation is a little confusing. you can improve the reader's experience by improving this part.
Author Response
Please refer to the attached document
Author Response File: Author Response.pdf
Reviewer 3 Report
In this article, authors propose a through-the-wall human detection method based on commercial Wi-Fi device (TwSense). The authors also suggest mitigating the interference from wall materials and other environmental factors and utilizing the robust principal component analysis (OR-PCA) method to extract the target signal from Channel State Information (CSI). A clustering algorithm (K-means) is used to segment the Doppler shift images based on the motions. They analyze four material types of walls: Concrete, Gypsum, Wooden Doors, and Glass. Authors propose this method as emergency rescue and health monitoring of elderly people. Authors have claimed that this method is quite robust.
Authors have produced results with reference to various distances as well as for different postures of persons using an experimental setup. They have made a comparison of various methods of detection as well.
1. Although an extensive analysis has been presented, it would be a good addition if the authors present the convergence of the final detection by the proposed algorithm, in other words, how quick or how complex is the proposed algorithm in producing a final detection decision.
2. Using feature extraction of images, can mobility detection be improved or can people be dynamically detected?
3. Since the results show a natural trend of reduced accuracy as the distance between wifi transmitters and receivers increase, the authors have used fixed position for these. Can they suggest an optimal positioning of the routers for best and most accurate results?
4. Have the authors proposed a GUI for the suggested applications, that may be useful for non-technical users?
5. Results show most accurate results for glass and the worst for concrete, whereas most walls are of concrete, can the authors suggest enhanced detection technique for concrete based facilities?
6. Can the authors analyze the detection for greater distances than 3m? As expected the accuracy would reduce, however, the greatest distance at which there is a probability of 0.5 can be presented as well.
7. Several new references relevant to this research work are available in literature, it would be best to add these to the article and claim the novelty of work in this case.
Major english revision is required, at times there are lengthy sentences that lose their meaning and need to be shortened. The word different is repeated multiple times in same sentence, such as "This paper evaluates and verifies the robustness of the proposed system by analyzing the effects of different environments, different people, different distances, and using different wall materials".
Another example, "From the figure, we can see that the Doppler shift changes with the action and the Doppler shift caused by different actions are different, and the different Doppler effects caused by different actions at the same time are the factors that distinguish the presence of the human body whether or not".
The word different is used 66 times in the manuscript, authors need to improve the readability of the paper by rephrasing.
Author Response
Please refer to the attached document
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authors have addressed all my comments satisfactory
The authors have addressed all my comments satisfactory
Reviewer 3 Report
Authors have satisfactorily addressed the issues raised in the review.
The quality is improved from the first manuscript.
