SitPAA: Sitting Posture and Action Recognition Using Acoustic Sensing

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper describes a sitting posture and actions recognition technique using acousting sensing. The authors argue that this method of sensing have some advantages when compared to other methods, including the use of wearable sensors, computer vision and RF methods. 

1. The introduction and Related Work section are complete, but the authors would do well in providing not only qualitative but also quantitative data on the works they review. That is, the authors should report quantitative accuracy values of the references. 

2. The authors report a number of experiments to validate their proposed method, including experiments on the effect of clothing, distance and the physical attributes of the test subjects. However, there is nothing mentioned in the paper about differences of test equipment. Were all the experiments done using the same laptop computer? What would be the effect of using other devices, as they could have different hardware performance and even a different distribution of microphone/speakers in the body of the device.

3. The authors should include in their discussion of the results, a comparison table with previous works. The table could include references among the other techniques (RF, computer vision, wearables) but mainly previous works using the acoustic sensing method. The table should indicate what type of posture or action is recognized, and highlight the references that perform the recognition of the same type of posture/action that the manuscript is focused on.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The study titled "SitPAA: Sitting Postures and Actions Recognition using Acoustic Sensing" introduces a technique that use acoustic sensing to identify and classify sitting postures and actions. The system leverages Frequency Modulated Continuous Wave (FMCW) signals for transmitting and receiving sound waves. It then applies signal processing techniques to extract characteristics and categorize postures and activities. According to the authors, their method attains a notable level of accuracy and demonstrates resilience in the face of noise and variations in lighting conditions.

 

The paper's notable merits lie in its utilization of acoustic sensors, enabling precise distance measurement, and its thorough signal processing pipeline, which adeptly manages noise and recovers pertinent information. The approach exhibits exceptional accuracy in both posture and action identification, with an average accuracy of 92.08% for posture recognition and 95.1% for action recognition.

 

Nevertheless, the paper exhibits other shortcomings, such as inadequate assessment, suboptimal positioning of speakers and microphones, significant privacy implications, and restricted commercial feasibility. Moreover, the approach requires significant processing resources and only assesses a limited set of identifiable poses and movements. The signal processing pipeline suggested in the research is characterized by a high level of computing intensity, which may impose restrictions on its practical feasibility in various settings.

 

In order to enhance the study, the authors could broaden the scope of postures and actions, decrease computational intricacy, assess the applicability of the method to diverse environments, examine techniques for minimizing interference from other acoustic sources, and explore approaches for implementing the method on more widely accessible hardware. To enhance their contributions and increase the practicality and applicability of the suggested method in real-world scenarios, the authors should consider implementing these ideas.

Comments on the Quality of English Language

The English language used in the work is generally of high quality. Sentences are well-structured, grammar is correct, and vocabulary is appropriate. However, the language could be enhanced in a few spots.

Clarity: By utilizing simpler language and minimizing terminology, the paper could be made clearer.

Conciseness: The paper could be made shorter by removing extraneous words and phrases. The authors may, for example, use the phrase "the method" rather than "the proposed SitPAA method."

Consistency: The paper could be more consistent if the same terminology was used throughout. The authors may, for example, use the phrase "sitting posture" instead of "sitting position" and "sitting orientation."

Overall, the English language utilized in the work is of high quality. The study is simple to read, and the authors' point is obvious. The language could be improved with a few minor tweaks.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf