Key Points’ Location in Infrared Images of the Human Body Based on Mscf-ResNet

Round 1

Reviewer 1 Report

Please add more information about the infrared (IR)database used.  From where it was obtained, do you have bioethics authorization?  , body position with respect to the IR camera, the IR camera spectral band,  IR imaging calibration methods used, any IR image post-processing used before applying them. 

In the experimental analysis section you mention that  12 key points are used How did you select them.    In the same section,  you compare the key point right to the key point left,   why do that ?. Later you compare the predicted key point with the correct result, from where did you get the correct result.

Author Response

Dear Editors and Reviewers:

Thank you for your letter and for the reviewers’ comments concerning our manuscript entitled “Key Points Location in Infrared Image of Human Body Based on Mscf-ResNet” (ID: futureinternet-1493958). Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval. Revised portion are marked in red in the paper. The main corrections in the paper and the responds to the reviewer’s comments are as flowing:

Question A: Please add more information about the infrared (IR)database used.  From where it was obtained, do you have bioethics authorization?  , body position with respect to the IR camera, the IR camera spectral band,  IR imaging calibration methods used, any IR image post-processing used before applying them. 

Answer A: Thanks for the helpful comment. I have added more information about the use of the data set in section 4.1.

(1) The data set is produced and collected by ourselves and have bioethics authorization.

(2) This experiment uses a short-wave infrared camera with a spectral response of 0.9um ~ 1.7um, which is calibrated using a calibration method based on active vision.

(3) The three postures (Stand with hands raised, stand akimbo, stand with hands down) of the front and back of the human body are photographed respectively.

(4) Preprocess the data set before using it. First, the data enhancement technology is used to preprocess the infrared thermal image of the human body, and then made it into a data set with a resolution of 1280 × 720.

Question B: In the experimental analysis section you mention that  12 key points are used How did you select them.    In the same section,  you compare the key point right to the key point left,   why do that ?. Later you compare the predicted key point with the correct result, from where did you get the correct result.

Answer B: Thanks for the helpful comment.

(1) The 12 key points are chosen because these 12 points represent the main joint points of the human body, and these joint points can well reflect the movements of the human body. It is the most significant point in human body pose estimation.

(2) We located the left and right key points of the same part, and made statistics on the accuracy of each point. In fact, the left and right key points on the same part do not matter, because the accuracy of these two shutdown key points will be affected with the difference of the posture. Therefore, each point is independent, and we only focus on the accuracy of their positioning.

(3) We compare the predicted value with the true value. The predicted value is generated by the network model, and the true value is artificially annotated by us. We use human eyes to observe and then mark the points to get the true coordinates of the key points. Although there are minor errors, the true value of the overall data is accurate. Therefore, there will be no impact.

Reviewer 2 Report

The manuscript propose a multi-scale convolution fusion deep residual network (Mscf-ResNet) model for human body infrared image positioning.
Results show that the proposed method has higher key-point positioning accuracy than other methods.
I find the topic interesting and being worth of investigation and the document is well strucutred, organized, fluidly written, has enough background information, the methodology followed is clearly explaineda, the results are clearly presented and support the conclusions.
Although I propose the following suggestions:
- I strongly suggest authors from refraining using personal pronouns such as "we" and "our" throughout the text and I encourage them to write it in an impersonal form of writing.
- Abstract requires structuring such as: problem, motivation, aim, methodology, main results, further impact of those results.
- At abstract authors refer: "Temperature differences in human organs can reflect problems in human health", this is only true in internal temperature measurements, which is not the case of infrared thermography that only perceives skin surface temperature.
- There is no credible scientific evidence to date of infrared thermography accurately diagnose breast cancer, I strongly suggest authors to delete this information and consider replacing the images shown in figure 1 by other body part image.

Author Response

Dear Editors and Reviewers:

Thank you for your letter and for the reviewers’ comments concerning our manuscript entitled “Key Points Location in Infrared Image of Human Body Based on Mscf-ResNet” (ID: futureinternet-1493958). Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval. Revised portion are marked in red in the paper. The main corrections in the paper and the responds to the reviewer’s comments are as flowing:

Question 1: I strongly suggest authors from refraining using personal pronouns such as "we" and "our" throughout the text and I encourage them to write it in an impersonal form of writing.

Answer 1: Thanks for the helpful comment. I already remove the word “we” and “our” in the manuscript and rewrite them.

Question 2: Abstract requires structuring such as: problem, motivation, aim, methodology, main results, further impact of those results.

Answer 2: Thanks for the helpful comments. I have added and modified abstract. The improper expression in the first sentence of abstract has been modified. Increase the motivation of the research in this article, and the impact of the results obtained. The entire abstract is structured as follows: problem, motivation, aim, methodology, main results, further impact of those results.

Question 3: At abstract authors refer: "Temperature differences in human organs can reflect problems in human health", this is only true in internal temperature measurements, which is not the case of infrared thermography that only perceives skin surface temperature.

Answer 3: Thanks for the helpful comments. I agree with you.The difference in the temperature of human organs can show his health problems, but the infrared thermal image only shows the skin temperature, which is not enough to prove his health problems.

Therefore, I modified the abstract. The human body can generate infrared radiation, so infrared thermal images are more used in monitoring and tracking.

Question 4: There is no credible scientific evidence to date of infrared thermography accurately diagnose breast cancer, I strongly suggest authors to delete this information and consider replacing the images shown in figure 1 by other body part image.

Answer 4: Thanks for the helpful comments.I have deleted this picture and replaced it with a picture of the leg.

Related studies have shown that diseased legs can cause uneven temperature distribution. For example, blood vessels are blocked.

Round 2

Reviewer 1 Report

All the requested inquires have been adequately considered.