PoPu-Data: A Multilayered, Simultaneously Collected Lying Position Dataset
Round 1
Reviewer 1 Report
Major Comments:
- Abstract: The abstract provides a concise overview of the paper, but it lacks specific details about the dataset and its potential applications. It would be helpful to include some key findings or implications of the dataset to better engage readers.
- Introduction: The introduction provides a good background on lying posture detection and the importance of multimodal data. However, it would benefit from a more explicit statement of the research objectives and the motivation behind creating the PoPu-data dataset.
- Dataset Collection: The paper briefly mentions the process of collecting the dataset, but it lacks sufficient details on the data collection methodology. It is important to provide information on how the data was recorded, the hardware or sensors used, and any relevant details about the participants or data collection environment. This would enhance the reproducibility and reliability of the dataset.
- Dataset Description: The paper mentions that the PoPu-data dataset contains multimodal data, but it lacks a comprehensive description of the modalities included. Providing detailed information about the specific types of data collected (e.g., audio, video, accelerometer, etc.) and their respective formats would be beneficial.
- Dataset Size and Diversity: It would be helpful to include information on the size and diversity of the PoPu-data dataset. How many participants were involved? Were there any demographic factors taken into account? The inclusion of such details would allow researchers to understand the generalizability and applicability of the dataset to different populations.
- Annotation Process: The paper briefly mentions the annotation process, but it would be beneficial to elaborate on the methodology and criteria used for labeling the lying postures. How many annotators were involved? Was there an inter-rater reliability assessment conducted? Providing such details would ensure the reliability and consistency of the annotations.
- Results and Analysis: The paper does not include any results or analysis section, which is crucial for evaluating the dataset's quality and potential use cases. It would be valuable to present some initial findings or insights derived from the dataset, along with any statistical measures or evaluation metrics used.
- Error Analysis: Given that the dataset is likely to contain errors or noise, it would be helpful to include an error analysis section. This section could highlight the challenges or limitations of the dataset, potential sources of errors, and strategies for mitigating them.
- Grammar and Technical Suggestions:
- The paper contains several grammar and punctuation errors throughout. Proofreading and editing for clarity would greatly enhance the overall readability.
- Some technical terms or abbreviations are used without proper definitions or explanations. It is important to provide clear definitions or references for such terms to ensure understanding for readers not familiar with the specific field.
Suggestions:
- Provide a more detailed description of the dataset collection process, including hardware, sensors, and data recording details.
- Clearly describe the modalities included in the dataset, their formats, and any pre-processing steps applied.
- Include information on the dataset size, participant diversity, and any demographic factors considered.
- Elaborate on the annotation process, including the number of annotators, inter-rater reliability assessment, and annotation criteria.
- Present some initial results or analysis based on the dataset, including statistical measures or evaluation metrics used.
- Include an error analysis section to discuss potential dataset limitations and strategies for error mitigation.
- Conduct a thorough proofreading and editing pass to correct grammar, punctuation, and readability issues.
- Provide clear definitions or references for technical terms and abbreviations used in the paper.
By addressing these suggestions and making the necessary revisions, the paper can improve its clarity, reproducibility, and overall quality, thus enhancing the value and usability of the PoPu-data dataset.
Comments for author File: Comments.docx
Author Response
We would like to thank the reviewers for their thoughtful and valued comments. In the text (attached) we list all the reviews the manuscript received and try to answer the questions placed by the reviewer. Our answer, in red, is presented after each question, referencing, when appropriate, the manuscript section containing the alteration. All the alterations performed to the manuscript are also in red for easy identification by the reviewer.
Author Response File: Author Response.pdf
Reviewer 2 Report
The paper presents a novel set of data sets that can be used as a part of training a deep-learning system in order to determine the sleeping posture of a person using embedded pressure sensors.
Author Response
We would like to thank the reviewers for their thoughtful and valued comments. In the text (attached) we list all the reviews the manuscript received and try to answer the questions placed by the reviewer. Our answer, in red, is presented after each question, referencing, when appropriate, the manuscript section containing the alteration. All the alterations performed to the manuscript are also in red for easy identification by the reviewer.
Author Response File: Author Response.pdf