Adaptive Age Estimation towards Imbalanced Datasets

Round 1

Reviewer 1 Report

Dear authors, first of all, congratulations on the work. I leave some comments below.

1- In the abstract and the title, at first reading, it is not clear what technique is behind the age estimation (i.e. through the use of face images). I would clarify this better.

2- there are some writing errors in the text, it should be reread more carefully. For example at paragraph 175 'kernal' and at paragraph 197 'moethod', just to name a few. 

3- the idea of the method used in general is simply a method for handling highly unbalanced datasets. It should be made more complete. For example, it was chosen to proceed using CNN convolutional neural networks despite the fact that this is not a very innovative technique compared to the current state of the art. Despite this, it could also be fine if the method were more complete. For example: paragraph 246, why is a batch size of 64 and 150 fixed epochs used? What technique was used to select them? I suggest using a hyperparameter optimiser (e.g., Optuna) and setting an early stop on the number of epochs, thus allowing it to be raised as desired. The same applies to lambda parameters, learning rate, choice of optimiser (e.g., Adam) and all other hyperparameters involved. 

 

As it stands, it is not ready for publication, but once the necessary modifications have been made, after extensive reanalysis of the results, it can be evaluated again.

Should be improved.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

please refer the manuscript

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Dear Editor of Applied Sciences and Authors,

I have read the manuscript "Adaptive age estimation towards imbalanced datasets" and have found it scientifically sound and relevant, well-written (some typos and errors must be corrected, though) and warranting publication as is. I would only suggest to the authors to frame the "age estimation" from pictures in a more narrow scientific field, as "afe estimation" is a terminology used in a variety of fields that are not directly related to the contents of this manuscript.

Best regards.

Minor typos and errors must be revised.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thank you very much for the answers. 

Regarding the third comment then, if you don't want to do new tests and optimize the hyperparameters, I suggest you specify that in the methodology and conclusions, also in view of future developments of the method.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf