Model-Assisted Labeling and Self-Training for Label Noise Reduction in the Detection of Stains on Images of Laundry
Round 1
Reviewer 1 Report
all shortcuts need to be defined when used first time.. like CNN, etc.
overall, the paper is interesting. covering investigation on important topic.
Results are clear and promising.
unfortunately i recommend to improve discussion, conclusions and abstract.
In the discussion you need to refer to other journal sources and compare your results with other work by more compact way - at the end of section IV. The best is to add table where you will summarize your and other contribution to provide more reliable and clear discussion in one overview. Please don't repeat single evaluations done by single subsections in Section VI.
Also conclusion is shorter and not so strong. It need to include PROS and CONS of your solution in compare to other solutions or traditional ways with VALUES, %, just clear contribution in compact way.. It is also missing in abstract.
and future directions are missing and conclusion s also..
after incorporating these MINOR changes, article can be accepted
Author Response
Thanks for the nice review. We defined the shortcut CNN at its first position and scanned for others we might have missed. In addition, we added a table and wrote a short paragraph at the end of section IV detailing the difference of our approach to others. Furthermore, we re-structured the article by combining the sections "Results" and "Discussion" to avoid repetition. We also moved the future work section to the end of the conclusion and added a paragraph detailing advantages and disadvantages of our approach. The abstract and the conclusion now state the average improvement our approach yields in percentage.
Reviewer 2 Report
All figures and tables should be after being cited, for instance look at where you cited Figure 6, and where is it located.
“The better score of m0 and m1 is always” what “always” is doing here?
“Hence, the score of m1 being highlighted indicates a successful run. “ does not make any sense!
“Figure 6 shows RDR curves.”
I just went randomly over few parts, this manuscript is not readable and needs to be proofedited in depth. Do not just respond to my comments but review and rewrite your manuscript.
“e noisy or, in other words, may contain errors” makes no sense.
This is a math journal and there is not a single mathematical equation. I would like you to explain the algorithm in the math format so I could understand. Also prepare figures, explaining the process you make to reach your objectives, to do that you need also to highlight your problem statement or objectives.
Apparently going forward your paper the quality of manuscript gets poorer and poorer so pay especial attention to the end of your manuscript.
Again, rewrite your paper, organize it in a more academic way by including more equations and depicting the algorithms.
Answer my concerns point by point and highlight all changes!
Author Response
We made sure that all figures and tables are now displayed after being cited. We also had a native English speaker check the manuscript. In addition, we added Figure 5 to illustrate our procedure, and we included the equation for the measure we used.
Round 2
Reviewer 2 Report
addressed
