Analysis of HMAX Algorithm on Black Bar Image Dataset
Round 1
Reviewer 1 Report
In this study, a general and unspecific assessment of the classification capabilities of HMAX is presented. This paper addressed an important issue in the field, however, some information is missing and a revision is required. Please consider bellow comments to improve the quality of your manuscript.
The HMAX is not well defined and described in the abstract. What is HMAX? what does it stand for? why there is a need for the new dataset? why you did not use available standard datasets. Why HMAX is used and not other methods. what type of data included in the proposed dataset? All these questions should be answered in the abstract.
There are many other techniques for image classifications including deep networks, convolutional, and etc, why HMAX is considered? it should be discussed in the early introduction. Include a discussion about other models and why HMAX is considered?
Considering and evaluating HMAX seems to be not a significant contribution to this field. Moreover, you did not compare it comprehensively to several recent methods. Only CNN methods considered.
Based on Table 1, it seems CNN is better. Faster training (by HMAX) is not a significant advantage.
Classification error is important but other metrics, please consider including other metrics such as AUC, Precision, and Recall.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The study is about providing the Black Bar Image Dataset (BBID) and utilizing HMAX to classify the images. The results showed that the model achieved good effectiveness in several conditions. Overall, the dataset has potential interest to the reader, however, it is not enough to justify that the paper can be fit the scope of this Journal. Thus, the authors are suggested to submit it to the other Journal such as Data in MDPI. Otherwise, the major improvements should be carried out such as:
- This paper has huge contents; however, the purpose and contribution of the study is still unclear and vague.
- What is HMAX stand for? It should not be abbreviated in the abstract and the Introduction before presenting the full abbreviation of HMAX term.
- How the HMAX works? It should be illustrated in figure or pseudo-code form to give a better understanding for the reader.
- It is not enough to present one model (HMAX) in the study, the research paper should provide the comparison between more than one other model (from the state of the arts/ previous study)
- Other performance metrics should be presented such as sensitivity, specificity, accuracy, precision, recall, the area under the curve (AUC), and etc.
- Experimental setup as well as implementation details of the model should be revealed as detail as possible to promote reproducibility of the study.
- Overall, the main contents of the paper are about the BBID, thus the reviewer suggests the authors to submit it to other Journal such as Data instead of Electronics.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
All the comments have been addressed.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Thank you for addressing the reviewer concerns. However, the title of the paper should be changed if the purpose of the paper is to investigate or analyze the HMAX algorithm using Black Bar Image Dataset. The title could be changed to "Analysis of HMAX algorithm on Black Bar Image Dataset" or something else. In addition, if the comparison of different algorithms are not the scope of the study, it could be mentioned in the limitation or future study.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
