A Survey of Deep Learning-Based Human Activity Recognition in Radar

Round 1

Reviewer 1 Report

This work gives an extensive review on deep learning based human activity recognition in radar.  The topic is of particular interest and the content is comprehensive. There are some minor issues to be further improved: 1) More references could be involved in Section 3 to make this review more solid. 2) In Section 4, it is better to describe the deep learning methods used in the literatures in more detail,  including the data preprocessing, network design, loss function and other training tricks. 3) Figure 2(a) and 6 is blurred. A clearer one is desired. 4) Table 2 could be more compact in terms of contents.

Author Response

Many thanks for your precious time and efforts expended in an attempt to improve our
paper. Your insightful advice is very much appreciated. We attempted to address all your concerns,
which doubtlessly improved the paper and our hope is that our efforts will meet your approval. Our
detailed reflections are listed below point by point in the word file.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors present an interesting review of human activity recognition (HAR) based on different types of radars (Doppler, FMCW and UWB). I have small comments:

1. I have only a comment. I suggest to include in table 3 additional comments, for example,  the central frequency of the radar or the main application, or if not try to joint the rows for the same type of radar and deep model.

2.Another comment: I ignore if there is any result about the application of noise radar for HAR applications. 

Author Response

Dear reviewer,

        We feel quite grateful for the time and efforts you expended in an attempt to improve our
paper. Your insightful advice is very much appreciated. We revised our manuscript and attempted to address all your concerns. We hope that our efforts will meet your approval. Our detailed reflections are listed point by point in the word file.

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper provides an interesting overview of the deep learning approach to human activity recognition using radar data. It was well written and I enjoyed reading it. Below are some minor comments.

1. Fig. 5(a) needs to have a corresponding system diagram for clarity.

2. The work in Ref. 28 mainly provides time-range maps. Therefore, please check if the discussion about this reference in the paper (e.g., Fig. 2b, Table 3, line 284) is proper.

Author Response

Dear reviewer,

    Many thanks for your precious time and efforts expended in an attempt to improve our
paper. Your insightful advice is very much appreciated. We attempted to address all your concerns,
which doubtlessly improved the paper and our hope is that our efforts will meet your approval. Our
detailed reflections are listed point by point in the PDF file.

Author Response File: Author Response.pdf