Next Article in Journal
Early Season Forecasting of Corn Yield at Field Level from Multi-Source Satellite Time Series Data
Previous Article in Journal
A UAV-Based Single-Lens Stereoscopic Photography Method for Phenotyping the Architecture Traits of Orchard Trees
Previous Article in Special Issue
Radar Signal Classification with Multi-Frequency Multi-Scale Deformable Convolutional Networks and Attention Mechanisms
 
 
Article
Peer-Review Record

SCRP-Radar: Space-Aware Coordinate Representation for Human Pose Estimation Based on SISO UWB Radar

Remote Sens. 2024, 16(9), 1572; https://doi.org/10.3390/rs16091572
by Xiaolong Zhou, Tian Jin *, Yongpeng Dai, Yongping Song and Kemeng Li
Reviewer 1:
Reviewer 2: Anonymous
Reviewer 3: Anonymous
Remote Sens. 2024, 16(9), 1572; https://doi.org/10.3390/rs16091572
Submission received: 21 February 2024 / Revised: 24 April 2024 / Accepted: 25 April 2024 / Published: 28 April 2024
(This article belongs to the Special Issue State-of-the-Art and Future Developments: Short-Range Radar)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper introduces SCRK-Radar, a novel approach for human pose estimation (HPE) using SISO UWB radar. By reconceptualizing HPE into separate classification tasks for vertical and horizontal coordinates, the method enhances radar-based pose estimation's accuracy and robustness.

 

However, there are two obvious issues with this paper,

1. The English proficiency and readability of the paper are poor, especially for the naming of abbreviations, such as in the abstract section. "Space aware coordinate representation for human pose estimation Based on Single Input Single Output (SISO) Wideband (UWB) radar" is abbreviated as SCRK Radar, and what does the letter K represent? Widdeband is abbreviated as UWB, where does the letter U come from.

2. Innovation is unclear. Please further enhance the explanation of the innovative part, such as explaining the differences with existing methods, which improvements have been made, or which ones are original.

 

Other suggestions:

1. What is the meaning of Capital H in Figure 3, which needs to be explained in the paper or figure.

2. Explain how PCK is calculated in the Evaluation metric in Section 5.

3. What process was followed for collecting action data, and how were the micro-Doppler spectrograms designed? Was there any specific processing like full duration capture or selective feature extraction?

4. Could the authors compare SCRK-Radar with existing methods, noting how it differs or improves upon them?

5. The fonts in Figure 3 are inconsistent and need to be modified

6. What if there is obstruction to the object?

7. Figure 7 needs to clarify the description of time in the horizontal coordinate, not simply written as an abbreviation

8. Figure 8 obtains the spectrum of the motion of each part of the human body, the images description is incomplete, and the font is too small to read.

Comments on the Quality of English Language

The English proficiency and readability of the paper are poor, especially for the naming of abbreviations, such as in the abstract section. "Space aware coordinate representation for human pose estimation Based on Single Input Single Output (SISO) Wideband (UWB) radar" is abbreviated as SCRK Radar, and what does the letter K represent? Widdeband is abbreviated as UWB, where does the letter U come from.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Summary: The manuscript introduces SCRK-Radar, a novel approach for human pose estimation using SISO UWB radar. It distinguishes itself by conceptualizing pose estimation as two separate classification tasks for vertical and horizontal coordinates. The method utilizes radar echo signals processed to construct micro-Doppler (MD) matrices, further segmented for feature extraction. The proposed framework leverages HRNet and LiteHRNet architectures, demonstrating robust performance on the HPSUR dataset with an average error below 40mm across key-points of skeletal.

Strengths:

  1. Innovative approach to human pose estimation using radar data, addressing the challenge with a novel space-aware coordinate representation.

  2. Comprehensive dataset and experimental setup, providing a solid basis for evaluating the proposed method.

  3. Detailed methodology, including data preprocessing and feature extraction strategies, enhancing the potential for accurate pose estimation.

Questions:

  1. Lack of clarity regarding the definition and calculation of "normalized distance" in the evaluation metrics.

  2. Absence of direct comparison with existing methods, which limits the assessment of the proposed framework's advancements.

  3. Insufficient detail on the design and preprocessing of micro-Doppler spectrograms for action acquisition, crucial for understanding data handling and model training.

Minor Comments:

 1. Some sections could benefit from further elaboration, particularly on parameter selection and threshold settings for evaluation metrics.

 2. Enhancements in the comparative analysis section could significantly strengthen the paper's impact.

Recommendation: Considering the innovative approach and the potential impact of the SCRK-Radar framework on the field of human pose estimation, I recommend this paper for revision. Addressing the outlined weaknesses, particularly enhancing the comparative analysis and clarifying the methodological details, would substantially improve the paper's contribution to the literature.

Comments on the Quality of English Language

English writing need to be improved

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript:

“SCRK-Radar: Space-Aware Coordinate Representation for Human Pose Estimation Based on SISO UWB Radar”, by

X. Zhou, T. Jin, Y. Dai, Y. Song and K. Li (Ref. No.: remotesensing-2904977-peer-review-v1),

contains interesting material. However, it is not well-organized and requires a significant elaboration. In particular, the authors do not compare their method to other known methods. They also did not discuss the advantages and disadvantages of this method.

English is acceptable. However, some minor amendments may be desirable. The manuscript requires a few citations.

Abstract

1) The novelty or originality of this work should be reflected.

2) The key quantitative results, obtained in this study, should be shown.

1. Introduction

1) The sentence: “One of the most promising applications of short-range radar is in human pose estimation, where the technology is used to determine the position and orientation of a person's body parts”, should be cited.

2) The sentence: “Doppler spectrum, distinct from those generated by other 100 actions such as running or waving. This distinction arises because each type of movement 101 has a unique velocity profile over time, captured by the micro-Doppler effect”, should also be cited.

3) The sentence: “By adapting Simcc for radar-based human pose estimation, the SCRK-Radar method transforms the challenge of pinpointing human key points into two distinct classification tasks: one for vertical coordinates and another for horizontal coordinates”, requires more clarifications. In particular, it is not clear how Simcc for radar-based human pose estimation can be adapted. This should be briefly clarified.

2. Related Work

1) This section should be renamed as Related Works.

2) The sentence: “The representation of coordinates plays a crucial role in accurately modeling and predicting the positions of human joints and limbs”, should be cited.

3. Theory

1) Perhaps, the sections 2 and 3 should be merged together since this section represents the continuation of the section 2.

2) All equations that were not derived by authors, should be cited.

4. Method

1) The sentence: “This paper utilized a SISO UWB radar, specifically an FMCW radar operating within the 2.7 to 3.2 GHz range and featuring a bandwidth of 500MHz”. Why particularly this spectral ranger from 2.7 GHz to 3.2 GHz is applied? What is an advantage of this spectral range?

5. Results

1) What are the main advantages of the proposed method for the space-aware coordinate representation for human pose estimation as compared to other known methods?

2) What are the disadvantages of the proposed method?

The manuscript requires a major mandatory revision.

Comments on the Quality of English Language

Minor English corrections are required.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

 

The author has suggested changes according to the requirements and the paper can be published directly.

Author Response

Dear Reviewer:

Thank you for the comments regarding our manuscript. We are grateful for the reviewer’s insightful feedback and constructive suggestions.

We are pleased to hear that the reviewer has found our revisions satisfactory and that our paper can be published directly. We sincerely appreciate the guidance and expert suggestions provided throughout the review process, which have undoubtedly enhanced the quality and impact of our work.

Thank you once again to the thoughtful review and to you for overseeing this process. Please let us know if there are any further steps we need to take to facilitate the publication of our paper.

Best regards,

Zhouxiaolong

National University of Defense Technology

[email protected]

Manuscript ID: remotesensing-2904977

Back to TopTop