Deep Learning and Transfer Learning for Automatic Cell Counting in Microscope Images of Human Cancer Cell Lines
and Jakub M. Tomczak 1,*,†Round 1
Reviewer 1 Report
This is a well-written paper containing interesting results which merit publication. For the benefit of the reader, however, a number of points need clarifying require further justification. Those are given below.
Comments:
In Figure 2, photographs that show an image is unclear (especially the Right image is so dark to judge image processing).
In Figure 4, the authors should show an enlarged portion as “square” on the Left side of Figure 4.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This is the review report of the paper titled "Deep learning and transfer learning for automatic cell counting in microscope images of human cancer cell lines".
The paper is good. However, I would suggest some comments to improve the paper further.
1- The abstract has to be improved by mentioning the highest results achieved.
2- Most of the references are very old, enrich the paper with new ones.
For example, it was mentioned in the paper " More recently, Deep Neural Networks (DNN) have been applied for a wide range of computer vision tasks and has set new state-of-the-art on several benchmarks related to classification, object detection, and segmentation [29–31]"
Ref.29 and 31 are very old which can be replaced with the following references
https://link.springer.com/article/10.1186/s40537-021-00444-8
https://ieeexplore.ieee.org/abstract/document/8694781
3- Clarify the differences between this work and the following paper in terms of the methods, can this work compare to it in terms of results?
https://www.tandfonline.com/doi/full/10.1080/21681163.2016.1149104
4- Since there is no related work section. it is required to show the issues with previous methods that solved in this work.
5- The used dataset is very small, even with image augmentation techniques, plus ResNet is very deep, the results are high. prove the model is not overfitted.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
This is an interesting research paper. There are some suggestions for revision.
1. The motivation is not clear. Please specify the importance of the proposed solution.
2. As shown in introduction, the second contribution needs to have more descriptions. The current descriptions do not show the unique contribution clearly. The third contribution cannot be counted as a contribution. The comparative analysis is a type of verification.
3. Please discuss more recently published solutions, especially the solutions published in 2021 and 2020.
4. Section 2.2 shows several feature extractors and regressors. But it does not clearly show how to select a suitable extractor or regressor for a certain type of data.
5. Identity mapping and projection mapping need more explanations.
6. As shown in Fig. 3 (b), please explain how to achieve the adaptation.
7. Please discuss the special features of the data. Which designs of the proposed solution are specialized for the data?
8. The comparative experimental results are not convincing. Please compare the proposed solution with more recently published solutions.
9. It seems the proposed solution is a combination of several existing solutions. Please specify your contributions.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
The authors addressed the comments very well.
Good Job.
Reviewer 3 Report
All my concerns have been addressed. I recommend this paper for publication.
