Next Article in Journal
Mechanical Models for Comparative Analysis of Failure Characteristics and Groundwater Inrush of Coal Seam Floors
Next Article in Special Issue
An Efficient Boosting-Based Windows Malware Family Classification System Using Multi-Features Fusion
Previous Article in Journal
A Wideband Microstrip-to-Waveguide Transition Using E-Plane Probe with Parasitic Patch for W-Band Application
Previous Article in Special Issue
Robust Classification Model for Diabetic Retinopathy Based on the Contrastive Learning Method with a Convolutional Neural Network
 
 
Article
Peer-Review Record

An Efficient Attention-Based Convolutional Neural Network That Reduces the Effects of Spectral Variability for Hyperspectral Unmixing

Appl. Sci. 2022, 12(23), 12158; https://doi.org/10.3390/app122312158
by Baohua Jin, Yunfei Zhu, Wei Huang *, Qiqiang Chen and Sijia Li
Reviewer 1:
Reviewer 2:
Appl. Sci. 2022, 12(23), 12158; https://doi.org/10.3390/app122312158
Submission received: 26 October 2022 / Revised: 23 November 2022 / Accepted: 23 November 2022 / Published: 28 November 2022
(This article belongs to the Special Issue Advances and Application of Intelligent Video Surveillance System)

Round 1

Reviewer 1 Report

The authors have presented a novel CNN method to reduce spectral variability and improve the efficiency of. Hyperspectral unmixing. The paper is well written. The authors can introduce a discussion section wherein they can discuss the relevant applications of the proposed method and justify the novelty 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

In order to obtain accurate remote sensing hyperspectral feature information, the paper "Efficient Attention-Based Convolutional Neural Network with Reduced the Effects of Spectral Variability for Hyperspectral Unmixing" puts forward a kind of efficient attention-based convolutional neural network (EACNN) and an efficient convolution block attention module (ECBAM), which has certain practical application value.

However, the reviewer believes that there are some problems in this article, including but not limited to:

1. Line 93, Line 95 and Line 97: For the names (DEAN, TANET, AE) that appear for the first time in the paper, spell the full name, and then write the abbreviation.

2. Line 98: There are many kinds of end-to-end CNN, which should be clearly pointed out here.

3. Line 119,138:I think the author's main contribution should be more concise, which is a bit tedious here, so please author consider it.

4. The author's description of figure 2 is not accurate enough, and many of the detailed features of the network structure are not described, so please correct it.

5. Line 308-310, the author has not accurately described the source, type and size of the datasets. Please supplement it. For the deep learning method, the requirements for the size of the datasets are relatively high, and the number of datasets used by the author is not described in the article.

6.  In Table 3, please describe in detail what Material 1-5 represents.

 

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The author has revised the article and it can be accepted.

Back to TopTop