Review Reports

Round 1

Reviewer 1 Report

1 The meaning of IOU should be explained clearly. There is no explanation of IOU(.), so that the paper is difficult to understand.

2 The paper discribed that the EfficientNet series network has a total of 8 models of different  sizes, then why EfficientNet-B0 is chosen? The reason should be discribed.

3 A whole structure of the algorithm shoud be clearly presented. Fig 1 is not clear and it should show every methods discrbed in the following part of this paper.

Author Response

Thanks for the suggestion,Please see the attachment for your reply. 

Author Response File: Author Response.docx

Reviewer 2 Report

 

The authors provided quite a straightforward and well-organized manuscript on the actual topic. Nevertheless, I noticed some issues to be improved before the publication.

1.       The quality of some pictures is too low. The fonts in the figures are too small and not readable.

2.       The authors provided detection accuracy for the various objects, but it is not clear how it was calculated. Whether those objects were in every picture or were they randomly appearing. From my point of view, it is necessary to describe the methodology and experiment in a more detailed way.

3.       The reference overview is quite weak only 21 sources, some of them quite old. From my point of view, it is necessary to make a more detailed overview and highlight the novelty and motivation for such an approach.

4.       It would be nice if authors could discuss their method limitations and provide recommendations for future research.

 

Author Response

Thank you for  the suggestion,Please see the attachment for your reply.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Authors take into account all my remarks. I am happy with improvements.