Research on Retinex Algorithm Combining with Attention Mechanism for Image Enhancement
Round 1
Reviewer 1 Report
This paper proposes an improved Retinex Net algorithm for low illumination 8 image enhancement, which is based on Decom Net and Enhance Net structures of Retinex network. The work proposed by the authors is good and validated through the experimental results. Although I suggest the Authors that use "Proposed Method" isteed of the "Ours" and literature part need some improvement in final manuscript.
Author Response
Point 1: Authors are suggested that use "Proposed Method" instead of the "Ours" and the literature part needs some improvement in the final manuscript.
Response 1: We attach great importance to your suggestions and have carefully read the contents of the suggestions. We have corrected the grammar properly, optimized the wording, and replaced the first person expression like "Ours" in the article with "the proposed method" or "this paper".
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Proposed study is presenting a Novel method for improving of low light image. Results are presented in an organized manner. However, table3 so the benchmark results are not sufficient to my Side. More state of the art methods should be included, especially the methods based on deep networks. Furthermore, a grammatical and language check is needed. Since hazy images have the same poor contrast problem, the low light image enhancement methods based on dehazing should be discussed in the literature review part.
Author Response
Point 1: The benchmark results are not sufficient and more state of the art methods should be included, especially the methods based on deep networks.
Response 1: We attach great importance to your suggestions and have carefully read the contents of the suggestions. In the last experiment of the manuscript, we added two kinds of deep networks to compare with the methods in this manuscript, namely GLADNet and EnlightenGAN. The enhancement effect and objective evaluation results of the two added image enhancement methods are shown in Figure 8 and Table 3 in Section 4.3. In addition, because the experimental results of the LIME algorithm used in the original manuscript will significantly overexpose the enhanced image, it has been deleted in the revised manuscript.
Point 2: The grammatical and language check is needed.
Response 2: We attach great importance to your suggestions and have carefully read the contents of the suggestions. We have checked all the drawings and corrected the grammar properly, optimized the wording, and replaced the first person expression like “Ours” in the manuscript with “the proposed method” or “this paper”.
Point 3: Since hazy images have the same poor contrast problem, the low light image enhancement methods based on dehazing should be discussed in the literature review part.
Response 3: We attach great importance to your suggestions and have carefully read the contents of the suggestions. By reading the literature “Single Image Haze Removal Using Dark Channel Prior” and “Fast efficient algorithm for enhancement of low lighting video”, we study the relationship between low light image enhancement and defogging. Therefore, in the introduction of this paper, we add a description of the low light image enhancement based on the defogging algorithm mentioned in the manuscripts. The above two references have been added to the reference lists [6] and [7].
Author Response File: Author Response.pdf
Reviewer 3 Report
The authors revised the article very well. All the suggestions are incorporated for previews reviewer comments, the article is in good shape and addresses all concerns. This can be accepted further.
Author Response
Point 1: The article is in good shape and addresses all concerns. This can be accepted further.
Response 1: We attach great importance to your suggestions and have carefully read the contents of the suggestions. We are very grateful for your help and feel honored to have your approval of the manuscript.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
The authors has made good improvements on the manuscript. It is acceptable to my side.
