Off-Axis Holographic Interferometer with Ensemble Deep Learning for Biological Tissues Identification
Round 1
Reviewer 1 Report
No comments
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
This manuscript proposes the object classifier using the deep-learning method. More specifically, The digitally captured off-axis hologram data is shown to the classifier, and the target object is classified. There seems no novelty in the main goal of this manuscript and the previous study in Ref.[6] at first glance, but I appreciate that the experimental validation is performed in the present study, which is not provided in Ref.[6]. And the presented success rate is remarkable, even though the data is corrupted by some kind of noise. The quality of the manuscript as well as the successful demonstration of the proposed deep-learning-based classifier looks fair enough to be published in the current journal with some improvements. Please consider the comments below to further improve the manuscript.
1. What is the input data to the classifier? Fourier filtered complex field? Magnitude or phase? I cannot easily find what it is.
2. The reconstructed images in Fig. 4 and Fig. 6 are unrecognizable. Even though they might contain some level of noise, the image looks not normal. It looks saturated, and I think more detailed features can be revealed after modifying the grey level of the obtained reconstructed field.
3. It would be better to present spectrum frequency plots in 2d with color maps and a color bar. It looks messy and hard to identify which part is clipped from the full spectrum field.
4. The discussion section looks redundant.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
After correction and additions, the article has become more intelligible and can be published
Author Response
Response to Reviewer 1 Comments
Point 1: After correction and additions, the article has become more intelligible and can be published.
Response 1: We appreciate your careful review and your hardworking for the improvement of this manuscript.
Yanmin Zhu, H.H.S. Lam, and Prathan Buranasiri
