Next Article in Journal
Microfluidics-Based Four Fundamental Electronic Circuit Elements Resistor, Inductor, Capacitor and Memristor
Next Article in Special Issue
Dynamic Deep Forest: An Ensemble Classification Method for Network Intrusion Detection
Previous Article in Journal
Programming Protocol-Independent Packet Processors High-Level Programming (P4HLP): Towards Unified High-Level Programming for a Commodity Programmable Switch
Previous Article in Special Issue
A Deep Learning-Based Scatter Correction of Simulated X-ray Images
Open AccessArticle
Peer-Review Record

Ship Target Detection Algorithm Based on Improved Faster R-CNN

Electronics 2019, 8(9), 959; https://doi.org/10.3390/electronics8090959
Reviewer 1: Norifumi Kawabata
Reviewer 2: Anonymous
Electronics 2019, 8(9), 959; https://doi.org/10.3390/electronics8090959
Received: 25 July 2019 / Revised: 14 August 2019 / Accepted: 27 August 2019 / Published: 29 August 2019
(This article belongs to the Special Issue Deep Neural Networks and Their Applications)

Round 1

Reviewer 1 Report

Thank you for submitting your original paper “Ship Target Detection Algorithm Based on Improved Faster R-CNN”. In this time, it is hard to accept for publication because there are many questions by reviewer.

 

(a). It is too short to explain in “1. Introduction”. You should describe more.

(b). In this paper, what do you clarify and purpose?

(c). Particularly, why you use R-CNN in this paper? Do you also use other methods?

(d). For Fig.1, it is hard for readers to understand because I do not know what you would like to achieve. On the other hand, for Fig.2, it is easy for readers to understand.

(e). For Fig.3, you should show the difference of specification between original image and downscale image.

(f).  For Fig.4, 6, and 7, I do not know what you would like to achieve because numerical value is not added.

(g). For Fig.5, 8 and 11, you had better add numerical parameter.

(h). For Fig.9 and 10, the scale of two graphs is not same.

(i).  For Fig.14, it is too small to see result by numerical value.

(j).  Finally, what do you achieve for novelty and knowledge in this paper?

(k). You had better search references to discuss this topic in this paper.



Author Response

Reply in detail to see attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper presents a ship target detection algorithm based on improved faster R-CNN. Results show that the proposed method can significantly shorten the detection time of the algorithm and improve the detection accuracy of Faster R-CNN algorithm.

 

As a general comment, I think the paper makes an interesting contribution to the literature by providing a new ship target detection method. At the same time, I think the paper needs some improvements before to be published in this journal.

Broad comments

The English language must be improved, in particular the introduction and the conclusion sections. The introduction is too poor; the authors should introduce the readers to the paper. I suggest the authors to enrich the introduction, at least, mentioning the use of Neural Network with different optimization methods in different field, such as in the prediction of the electricity price. I suggest also to add the following citations, where in particular in the second there is a comparison among different tools: Rafał Weron,“Electricity price forecasting: A review of the state-of-the-art with a look into the future”, International Journal of Forecasting, Volume 30, Issue 4, Pages 1030–1081, 2014 Silvano Cincotti, Giulia Gallo, Linda Ponta, Marco Raberto, “Modelling and forecasting of electricity spot-prices: Computational intelligence vs classical econometrics”, AI Communications, Volume 27, Issue 3, Pages 301-314, 2014 Nima Amjady, Farshid Keynia, “Day ahead price forecasting of electricity markets by a mixed data model and hybrid forecast method”, International Journal of Electrical Power & Energy Systems, Volume 30, Pages 533–546, 2008 Section 3 should not start with an image, and also the name of section 3 should not be “Article method” but, for example, only “methodology”. Section 3.3 line 199 “Theme” is a subtitle? Please organize better the section. Subsection 4.2 should not start with a figure. Please put the figure inside the section. In figure 9, there is something written in Chinese, please translate in English. Moreover, x and y labels should be added to the plot. Please add a table with the main variables and relative values used in the experiments. For example, the length of the data, the number of data used to train the neural network, etc. Moreover is not clear the size of the training and validation period.

Are the results of the paper influenced by the training period?

In order to compare the results, I would suggest the authors to evaluate also the MAPE error.

 

 

Specific comments

Line 20: I suggest the author to modify “R-CNN; Furthermore” with “R-CNN. Furthermore”

Line 21: I suggest the author to modify “to optimizing” with “to optimize”

Lines 72 - 74: the sentence is not clear. Please modify.

Lines 110 - 111: the sentence is not clear. Please modify.

Lines 115 - 116: the sentence is not clear. Please modify.

Lines 122 - 126: the sentence is not clear. Please modify.

Lines 140 - 147: These sentences are the same of line 129 - 136. Please remove

In figure 12 I would suggest the authors to use not only different colour in order to plot the error comparison, but also different line style or marker so that also in black and white the figure is readable.

Author Response

Reply in detail to see attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The paper has been improved following the referees’ suggestions, and now, according to me, it is ready to be published in this journal.

Back to TopTop