A Multi-Target Detection Method Based on Improved U-Net for UWB MIMO Through-Wall Radar
, Di Zhao 1,2, Kun Yan 1,2, Jinliang Nie 1,2,*, Bin Zhou 1,2 and Guangyou Fang 1,2
Round 1
Reviewer 1 Report
1) I think the term ‘aliasing’ appeared many times in this paper should be ‘unresolving’.
2) The coordinates of x, y and z should be drawn in Figure 1.
3) Line 156, ‘…the improved Kirchhoff algorithm proposed in section 2.1’ should be ‘…the improved Kirchhoff algorithm introduced in section 2.1’.
4) The writing of abbreviations should be unified, such as ‘ReLU’ etc.
5) The text ‘ResNet Module’ at the bottom of Figure 4(a)(b) should be deleted.
6) Line 402 – 405, since the FLOPs and Params in the traditional U-Net network training are higher than the improved U-Net network, why is the training time actually shorter?
7) It is suggested to provide the results and analysis of the ablation experiment in Section 4 or Section 5.
8) I think the results in Figure 15 cannot be called ‘tracking result’. For multi-target tracking, you must explain what data association algorithm and filters are used. These results may be plots or measurements of targets.
Author Response
Dear Reviewer
We’d like to express our most sincere gratitude to the reviewer for your effort and patience in reviewing our manuscript. We deeply appreciate your constructive comments that greatly help improve the technical quality and the presentation of this manuscript.
The attached file is our point-by-point responses to the comments raised by the reviewers. The comments are reproduced and our responses are given directly afterward, the rewritten part is marked in red.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The authors of the paper proposed a method of detecting multiple targets through a wall by UWB radar, that is based on U-Net. I have no fundamental reservations about the research, which is well interpreted in the submitted paper. I consider the topic of the submitted contribution to be current. The research strategy and results are clearly described. I appreciate the proposal of the method of detection of multiple targets through the wall by UWB radar, that is based on U-Net. I consider the given network model to be correct. The simulation results correspond to the created network model. I have no fundamental comments on the mentioned procedures of the authors in the creation of the model and the results of the simulation. References are appropriate. I recommend publishing this post after slight modifications. At the end of the post, I recommend to evaluate the simulation results more thoroughly and compare them with other methods. It is not clear from the paper how the proposed network will react to other movement scenarios than those presented in the presented model. Therefore, at the end of the post, it would be appropriate to clarify how your proposed network will react to other configurations of the movement of people behind the wall than were mentioned in the proposed scenario.
No.
Author Response
Dear Reviewer
We’d like to express our most sincere gratitude to the reviewer for your effort and patience in reviewing our manuscript. We deeply appreciate your constructive comments that greatly help improve the technical quality and the presentation of this manuscript.
The attached file is our point-by-point responses to the comments raised by the reviewers. The comments are reproduced and our responses are given directly afterward, the rewritten part is marked in red.
Author Response File: Author Response.pdf
Reviewer 3 Report
The authors have selected an interesting research topic, and have proposed a study on detection and localization for short-range through-wall imaging systems. The paper is well prepared and reults are promising. However, the proposed work has lack novelty. There are some concerns;
My main concern is with the novelty of the design. The proposed U-net Method with combination of SE-Resnet modules seems not a unique method for this localization. A similar method is presented in [1]. In terms of performance, results could be sufficient. However, can the authors please address the uniqueness of the design in light of existing through-wall imaging systems or similar applications by comparing it with the previously published works?
How did you define the simulation model of the scenario? Is that a 3D propagation problem or any other MATLAB tool is used?
[1] http://dx.doi.org/10.1115/DETC2020-22675
Author Response
Dear Reviewer
We’d like to express our most sincere gratitude to the reviewer for your effort and patience in reviewing our manuscript. We deeply appreciate your constructive comments that greatly help improve the technical quality and the presentation of this manuscript.
The attached file is our point-by-point responses to the comments raised by the reviewers. The comments are reproduced and our responses are given directly afterward, the rewritten part is marked in red.
Author Response File: Author Response.pdf
