Semi-supervised atmospheric component learning in low-light image problem

Masud An Nur Islam Fahim; Nazmus Saqib; Ho Yub Jung

doi:10.1371/journal.pone.0282674

Peer Review History

Original SubmissionJanuary 3, 2023
17 Jan 2023 Decision Letter - Praveen Kumar Donta, Editor PONE-D-23-00220Semi-supervised atmospheric component learning in low-light image problemPLOS ONE Dear Dr. Jung, Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. Please submit your revised manuscript by Mar 03 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file. Please include the following items when submitting your revised manuscript: A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter. If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols. We look forward to receiving your revised manuscript. Kind regards, Praveen Kumar Donta, Ph.D. Academic Editor PLOS ONE Journal Requirements: When submitting your revision, we need you to address these additional requirements. 1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf 2. Please note that PLOS ONE has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, all author-generated code must be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse. 3. Thank you for stating the following in the Acknowledgments Section of your manuscript: "This study was supported by research fund from Chosun University, 2022." We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form. Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows: "This study was supported by research fund from Chosun University, 2022. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript." Please include your amended statements within your cover letter; we will change the online submission form on your behalf. 4. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide. 5. We note that Figure 7 includes an image of a [patient / participant / in the study]. As per the PLOS ONE policy (http://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research) on papers that include identifying, or potentially identifying, information, the individual(s) or parent(s)/guardian(s) must be informed of the terms of the PLOS open-access (CC-BY) license and provide specific permission for publication of these details under the terms of this license. Please download the Consent Form for Publication in a PLOS Journal (http://journals.plos.org/plosone/s/file?id=8ce6/plos-consent-form-english.pdf). The signed consent form should not be submitted with the manuscript, but should be securely filed in the individual's case notes. Please amend the methods section and ethics statement of the manuscript to explicitly state that the patient/participant has provided consent for publication: “The individual in this manuscript has given written informed consent (as outlined in PLOS consent form) to publish these case details”. If you are unable to obtain consent from the subject of the photograph, you will need to remove the figure and any other textual identifying information or case descriptions for this individual. [Note: HTML markup is below. Please do not edit.] Reviewers' comments: Reviewer's Responses to Questions Comments to the Author 1. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. Reviewer #1: Yes Reviewer #2: Yes ******** 2. Has the statistical analysis been performed appropriately and rigorously? Reviewer #1: Yes Reviewer #2: Yes ****** 3. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified. Reviewer #1: No Reviewer #2: Yes ****** 4. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here. Reviewer #1: Yes Reviewer #2: Yes ****** 5. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters) Reviewer #1: In this work, a hybrid of the classical atmospheric optimization approach and recent learning-based studies is proposed to address the low-light image enhancement problem, which is expected to bridge the benefits of two genres of techniques. 1, It should be explicitly noted in the manuscript that a lower NIQE score indicates a higher perceptual image quality. 2, This paper randomly select 500 images from the GLADNet dataset to start the model training, and then adopt the proposed semi-supervised learning scheme to train the model. A natural question is that why not direclty use all 5000 images with ground-truth to train the model using standard supervised learning? What's the advatange of the proposed semi-supervised learning over the supervised learning? 3, A mixed loss function is adopted to train the model. There should an ablation study to verify the contribution of each single loss term. 4, In addition to the adopted performance measure, the authors are recommended to also use UNIQUE[1], a SOTA NR-IQA metric, to validate the performance of competting methods. [1] Zhang, W., Ma, K., Zhai, G., & Yang, X. (2021). Uncertainty-aware blind image quality assessment in the laboratory and wild. IEEE Transactions on Image Processing, 30, 3474-3486. Reviewer #2: This paper proposes an end-to-end low-light enhancement network model based on the fact that the inverse and fogged images of low-light images have similar statistical information, and a semi-supervised learning strategy for tuning low-light image enhancement using the no-reference image quality evaluation metric (NIQE). The paper presents qualitative and quantitative comparisons with current mainstream methods on several common publicly available low-light datasets. Strengths 1. The proposed semi-supervised training strategy is relatively novel and the obtained results also achieve optimal and suboptimal performance on multiple datasets. 2. The trained network model produces smaller inference bias compared to other methods which is demonstrated in Section 4.3. Weaknesses 1. The specific structure of the network model is not described in detail. The physical model is from previous work LIME, which is slightly less innovative on network model structure. 2. Is it problematic to use the NIQE metric as a quantitative metric for comparison with other methods while the training strategy already uses it? Despite the small amount of explanation provided by the authors in the paper, more discussion on the rationality of the approach is still needed. ****** 6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy. Reviewer #1: No Reviewer #2: No ******** [NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.] While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step. https://doi.org/10.1371/journal.pone.0282674.r001
Revision 1
9 Feb 2023 Author Response We have the same response to reviews in pdf format in the submission package. The pdf format is stylistically organized better, and easier to read. We thank the reviewers for their kind critiques. We hope we were able to answer all of the concerns to satisfaction. We also want the reviewers to know that we will make all the codes and image results available upon the publication of the paper. Reviewer 1: It should be explicitly noted in the manuscript that a lower NIQE score indicates better perceptual quality. Response: In 3.1, we added that NIQUE quantifies the deviation from the image quality. We hope we made the paper more clear. We also noted in the evaluation section that a lower value is better for all three evaluation metrics. Reviewer 1: This paper randomly selects 500 images from the GLADNet dataset to start the model training and then adopt the proposed semi-supervised learning scheme to train the model. The natural question is why not directly use all 5000 images with ground truth to train the model using standard supervised learning? What's the advantage of the proposed semi-supervised learning over supervised learning? Response: We agree and understand that using more data with ground-truth will almost always produce a better result. The advantage of semi-supervised learning is that it requires fewer number of images with ground-truths compared to direct supervised learning. And ground-truths are not always easy to obtain. Thus, we hoped that one of the contribution is to show that even with smaller number of ground-truths, we can produce state-of-the-arts results. Essentially, however, the difference between using 500 and 5000 ground-truths were sufficiently small enough to give a merit to semi-supervised learning, where we can make an argument for an additional contribution. Reviewer 1: A mixed loss function is adopted to train the model. There should be an ablation study to verify the contribution of each single loss term. Response: Fig. 3 shows differences of SSIM and PSNR for each single loss term as they are being trained. Instead of using no-reference metrics, we did our ablation test using SSIM and PSNR from ground-truths. Interestingly enough, a single SSIM loss does not produce highest SSIM score in the ablation test images. This is probably due to lack of sufficient diversity or number of training data, which our approach can compensate through the physical model and mixture of different loss functions such as smoothness and brightness functions. Fig. 4 also shows qualitative assessment of different weight values in the mixed loss functions. We understand that more can be done to make our ablation study more complete, however, we hope the study presented here are of sufficient interest for publication. Reviewer 1: In addition to the adopted performance measure, the authors are recommended to also use UNIQUE [Zhang et al. 2021], a SOTA NR-IQA metric, to validate the performance of competing methods. Response: Thank you for referring to UNIQUE [Zhang et al. 2021], a good study on blind image quality assessment. In our study, we have cited this paper. At present, unfortunately, we are not able to include a full-fledged comparison of the UNIQUE metric for all of the datasets and compared studies. We have observed UNIQUE's wide acceptance in the BIQA studies, but for image restoration studies, this metric is comparatively new to the community. In our future study, we will investigate the prominence of the UNIQUE metric over restorations task from different domains and aim to present a versatile assessment. In the meantime, we hope our current manuscript with three different blind evaluation metrics is sufficient to present the merit of our study. Reviewer 2: The specific structure of the network model is not described in detail. The physical model is from previous work LIME [Guo et al. 2017], which is slightly less innovative on network model structure. Response: We use DnCNN [Zhang et al. 2017] for our network. We have updated the Fig.2 and the manuscript to describe the model. However, unlike many of the previous studies, we want to emphasize that structure of the network is not important to the proposed approach. In Fig. 2, our feature map aggregator is an universal block, which can be any network with a slight modification (One input, two outputs instead of one). Hence, any one of networks like ResNet/DnCNN/UNet [He et al 2016, Zhang et al. 2017, Ronneberger et al. 2015], etc. can be used as the core feature extractor, which learns the approximate features for $A(x)$ and $t(x)$ to solve equation (4) for low-light enhancement. Reviewer 2: Is it problematic to use the NIQE metric as a quantitative metric for comparison with other methods while the training strategy already uses it? Despite the small amount of explanation provided by the authors in the paper, more discussion on the rationality of the approach is still needed. Response: We understand the reviewer's concern, however, it is common for image restoration research to use the same metric for both training and testing. For example, typical super-resolution networks use SSIM/PSNR as the regularizer/loss function and the main comparison metric [Galindo and Pedrin 2019]. Additionally, our method does not provide network gradients to the neural network with respect to NIQE. NIQE merely acts as a selector for the next training batch. Indeed, an implicit bias from NIQE may be present, but the bias should not be great as methods that provide training gradient directly from test/training metric. We are aware of the fidelity concern for the no-reference metrics, especially NIQE, and yet, among regular no-reference metrics, we have observed NIQE is the most used by the researchers [Banik et al 2018, Zhang et al. 2022]. Therefore, by following our peers, we have used NIQE as a selector during the training and one of the validation metrics. We added more details on why we choose NIQE in section 3.1. References Partha Pratim Banik, Rappy Saha, and Ki-Doo Kim. Contrast enhancement of low-light image using histogram equalization and illumination adjustment. In 2018 International Conference on Electronics, Information, and Communication (ICEIC), pages 1–4, 2018. Eldrey Galindo and Helio Pedrini. Image super-resolution improved by edge information. In 2019 IEEE International Conference on Systems, Man and Cybernetics (SMC), pages 3383–3389, 2019. Xiaojie Guo, Yu Li, and Haibin Ling. Lime: Low-light image enhancement via illumination map estimation. IEEE Transactions on Image Processing, 26(2):982–993, 2017. Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. pages 770–778, 06 2016. Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III 18, pages 234–241. Springer, 2015. Kai Zhang, Wangmeng Zuo, Yunjin Chen, Deyu Meng, and Lei Zhang. Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising. IEEE transactions on image processing, 26(7):3142–3155, 2017. Weixia Zhang, Kede Ma, Guangtao Zhai, and Xiaokang Yang. Uncertainty-aware blind image quality assessment in the laboratory and wild. IEEE Transactions on Image Processing, 30:3474–3486, 2021. Zhao Zhang, Huan Zheng, Richang Hong, Mingliang Xu, Shuicheng Yan, and Meng Wang. Deep color consistent network for low-light image enhancement. In 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 1889–1898, 2022. Attachments Attachment Submitted filename: 20230209_response_to_reviews.pdf https://doi.org/10.1371/journal.pone.0282674.r002
20 Feb 2023 Decision Letter - Praveen Kumar Donta, Editor Semi-supervised atmospheric component learning in low-light image problem PONE-D-23-00220R1 Dear Dr. Jung, We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements. Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication. An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org. If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org. Kind regards, Praveen Kumar Donta, Ph.D. Academic Editor PLOS ONE Additional Editor Comments (optional): Reviewers' comments: Reviewer's Responses to Questions Comments to the Author 1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation. Reviewer #1: All comments have been addressed Reviewer #3: All comments have been addressed ******** 2. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. Reviewer #1: Yes Reviewer #3: Yes ****** 3. Has the statistical analysis been performed appropriately and rigorously? Reviewer #1: Yes Reviewer #3: Yes ****** 4. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified. Reviewer #1: No Reviewer #3: Yes ****** 5. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here. Reviewer #1: Yes Reviewer #3: Yes ****** 6. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters) Reviewer #1: (No Response) Reviewer #3: Paper Title: Semi-supervised atmospheric component learning in low-light image problem The paper is accepted in present form. All the comments raised are addressed well by the authors. ****** 7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy. Reviewer #1: No Reviewer #3: No ******** https://doi.org/10.1371/journal.pone.0282674.r003
Formally Accepted
28 Feb 2023 Acceptance Letter - Praveen Kumar Donta, Editor PONE-D-23-00220R1 Semi-supervised atmospheric component learning in low-light image problem Dear Dr. Jung: I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department. If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org. If we can help with anything else, please email us at plosone@plos.org. Thank you for submitting your work to PLOS ONE and supporting open access. Kind regards, PLOS ONE Editorial Office Staff on behalf of Dr. Praveen Kumar Donta Academic Editor PLOS ONE https://doi.org/10.1371/journal.pone.0282674.r004

Open letter on the publication of peer review reports

PLOS recognizes the benefits of transparency in the peer review process. Therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. Reviewers remain anonymous, unless they choose to reveal their names.

We encourage other journals to join us in this initiative. We hope that our action inspires the community, including researchers, research funders, and research institutions, to recognize the benefits of published peer review reports for all parts of the research system.

Learn more at ASAPbio .