Feb. 12, 2024
line 42 and 43 where is stated " All the recordings were stored locally on our server, thus there are no privacy concerns about this dataset" :
Although the reviewer understands the point of view of the researchers and the experimental setup presented, is mandatory to discuss privacy concerns from the point of view of a user that may pass in a room where this technology may be installed , and has no access to the local data server. For instance a room n a public building. In 2024 this is a critical concern that any researcher must include in the discussion of their work to be published.
I can give to the authors one of the major concerns found nowadays, the ability of someone DIY this solution for their busines using the presented research work after searching on google.
=========================== Reply to the Editor ==================
My review raises the privacy & safety concern as I've been mentioning in previous reviews made. This last one summarizes clearly my major concern considering the current, and future state, of science availability on the internet for anyone and everyone to use. This is now a responsibility every scientific researcher must have when publishing their work and findings.
In this last review of the document, authors mention privacy and safety from the authors point of view. nothing is said or mentioned about privacy and safety when someone on the internet buys the paper to use it for personal use. My query is about responsible science and ethics, and since this work is directly related to human presence detection I can enumerate a list of dangers and ill usages to do harm and even kill someone (terrorrim).
2023-08-21
This reviewer finds the theme and subject of this work to be of great interest and relevance to the Journal "Engineering Research Express". However, before proceeding to the next stage of acceptance some changes and modifications are needed. The recommendation status is "Unsatisfactory revision"
=========================================================
In this reviewer's opinion, is missing in this experimental research work (from the previous revision):
- a custom purpose-made hardware electronics (as opposed to using off-the-shelf PCB modules (not mandatory, but highly recommended) with its own dedicated section on the document. Not mandatory but highly recommended.
DONE - a description of the hardware electronics specifications
NOT DONE - a description of how a dataset is handled and stored. Is it on the proposed hardware electronics? Is it Cloud-based, for instance using an HTTP API for data exchange? A subsection is needed to describe this topic of "data storage and handling". (mandatory)
NOT DONE - This kind of technology handles sensitivity data, meaning data that can be put to criminal or even worst kinds of activities. This should be presented and briefly discussed in its own section on this research document. This reviewer suggests the inclusion of the topic "open data" and its advantages as a safer way to handle this kind of sensitive data. (mandatory)
- A link to an experimental data repository. For instance a dataverse, Figshare, dryad, or any other. Where experimental data can be evaluated by a reviewer. Not mandatory, but highly recommended.
- A link to a repository where the firmware code used can be evaluated by a reviewer. For instance on GitHub. Not mandatory, but highly recommended.
DONE - this reviewer highly recommends the inclusion of a pseudo-algorithm in the corresponding section of the document.
DONE - Where needed, is missing proper references of figures and tables on the text that is describing them. For instance, section 4.2 is missing a reference to Table 3 next to "...lead to overfitting".
DONE - "Table 5 - Comparison between different algorithms" must include an introductory paragraph to describe each algorithm presented in Table 5 and include all hyperparameters for each that led to results shown in Table 5.
ALMOST DONE - Grammar and text errors are now almost all corrected (those this reviewer was able to identify), so for this revision, R1 is OK.
- Due to the current status and evaluation proposed, this reviewer did not do ...
- ...manual verification of the references listed.
- mathematical formulation looks solid, however, this reviewer did not do an extensive verification.
This and all the above will change on future resubmissions and re-revisions of the document. When all identified tasks above are addressed. And added. In particular, those marked as "NOT DONE" this review consider of extreme importance given the current state of availability of smart devices to a common citizen on the internet, in particular, authors need to provide some considerations about safety, misusage, and privacy for the proposed smart device solution in a section/subsection of this document. See the items marked as "NOT DONE" for more information about what this reviewer is requesting.
===============================================================
July, 10 2023
This reviewer finds the theme and subject of this work to be of great interest and relevance to the Journal "Engineering Research Express". However, before proceeding to the next stage of acceptance some changes and modifications are needed. The recommendation status is "Major revision required before acceptance"
=========================================================
In this reviewer's opinion, is missing in this experimental research work:
- a custom purpose-made hardware electronics (as opposed to using off-the-shelf PCB modules (not mandatory, but highly recommended) with its own dedicated section on the document.
- a description of the hardware electronics specifications
- a description of how a dataset is handled and stored. Is it on the proposed hardware electronics? Is it Cloud-based, for instance using an HTTP API for data exchange? A subsection is needed to describe this topic of "data storage and handling". (mandatory)
- This kind of technology handles sensitivity data, meaning data that can be put to criminal or even worst kinds of activities. This should be presented and briefly discussed in its own section on this research document. This reviewer suggests the inclusion of the topic "open data" and its advantages as a safer way to handle this kind of sensitive data. (mandatory)
- A link to an experimental data repository. For instance a dataverse, Figshare, dryad, or any other. Where experimental data can be evaluated by a reviewer. (not mandatory, but highly recommended)
- A link to a repository where the firmware code used can be evaluated by a reviewer. For instance on GitHub. (not mandatory, but highly recommended) Also, this reviewer highly recommends the inclusion of a pseudo-algorithm in the corresponding section of the document.
- All acronyms used throughout the research document must first be presented and briefly explained. For instance, LVQ, GLVQ, GMLVQ, PIR, and all others are used.
- Where needed, is missing proper references of figures and tables on the text that is describing them. For instance, section 4.2 is missing a reference to Table 3 next to "...lead to overfitting".
- "Table 5 - Comparison between different algorithms" must include an introductory paragraph to describe each algorithm presented in Table 5 and include all hyperparameters for each that led to results shown in Table 5.
Due to the current status and evaluation proposed, this reviewer did not do ...
- ...manual verification of the references listed.
- ...extensive grammar or text error checking. However, on successive reads of the document, it was possible to identify a few errors (marked in red on the revised document sent) that need attention and correction. For instance in the legend of Figure 3, "movement of codebooks" needs correction as a codebook does not have the ability to move from one place to another. Only humans. In Figure 2 the "REC" legend is inadequate and should be replaced by, for instance, a datalogger or other that better describes the hardware electronics and its function.
- mathematical formulation looks solid, however, this reviewer did not do an extensive verification.
This and all the above will change on future resubmissions and re-revisions of the document. When all identified tasks above are addressed. And added. In particular, those marked as "mandatory".