4-Class MI-EEG Signal Generation and Recognition with CVAE-GAN
Round 1
Reviewer 1 Report
The manuscript has been well revised.
Author Response
Response to Reviewer 1 Comments
Point 1: The manuscript has been well revised.
Response 1: Great thanks for your suggestion
Reviewer 2 Report
The paper was improved following the suggestions of the reviewers. I have no further comments.
Author Response
Response to Reviewer 2 Comments
Point 1: The paper was improved following the suggestions of the reviewers. I have no further comments.
Response 1: Thank you again for your instructive criticism and correction.
Reviewer 3 Report
In the Introduction I would be glad to see a short paragraph introducing the BCI and its wide applications. I would be happy to see more references to most up-to-date interesting papers in the field, such as paper, Speed control of Festo Robotino mobile robot using NeuroSky MindWave EEG headset based brain-computer interface paper A Brain–Computer Interface Project Applied in Computer Engineering, paper Electroencephalogram-based brain-computer interface for internet of robotic things, and paper Examining the learning efficiency by a brain-computer interface system, etc.
Author Response
Response to Reviewer 3 Comments
Point 1: In the Introduction I would be glad to see a short paragraph introducing the BCI and its wide applications. I would be happy to see more references to most up-to-date interesting papers in the field, such as paper, Speed control of Festo Robotino mobile robot using NeuroSky MindWave EEG headset based brain-computer interface paper A Brain–Computer Interface Project Applied in Computer Engineering, paper Electroencephalogram-based brain-computer interface for internet of robotic things, and paper Examining the learning efficiency by a brain-computer interface system, etc.
Response 1: Thank you again for your instructive proposal on the structure of the paper.
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
I enjoyed reading the paper. The authors have presented an interesting article about 4-class MI-EEG Signal Generation and Recognition with CVAE-GAN.
How do you think your findings can be useful applicable to other engineering disciplines?
Introduction is clear, but please explain in more detail in this chapter how the article relates to these topics:
- Speed control of Festo Robotino mobile robot using NeuroSky MindWave EEG headset based brain-computer interface
- A Brain–Computer Interface Project Applied in Computer Engineering
- Electroencephalogram-based brain-computer interface for internet of robotic things
- Examining the learning efficiency by a brain-computer interface system
Reviewer 2 Report
The paper aims at proposing a method for the generation of simulated EEG signals, starting from Motor Imagery data, using a generative adversarial network in combination with a conditional variational auto-encoder approach.
The paper is quite hard to read and an extensive editing of English language and style is required. It is not clear the novelty of the proposed method and its applications into the practice are not understandable or missing.
The methods are confusing and lack a fundamental part related to the metrics used to compare real and simulated data.
The results don't convincigly show the robustness of the approach. For example in Figure 2 and Figure 4, strong differences are shown between real and simulated data. ERD and ERS appear really different. How can the authors support the goodness of their method if the simulated data are so different from the training dataset?
The metrics listed in table 4 don't help to understand the performance of the approach since detailed methodological information lack in the Materials and methods section.
The results don't support the conclusions.
Discussion is completely missing.
I believe the paper needs deep re-design work before being considered for publication.
Reviewer 3 Report
In this paper, the authors combined CVAE (conditional variational auto-encoder network) and GAN (generative adversarial network) for generating artificial data of electroencephalogram (EEG). They compared the generated data with real data, and reported that the former showed averaged waveforms similarly to the latter. They concluded that their approach is a good method for generating high-quality artificial EEG data.
This is well-written and I have no objection to the contents at this moment. My only concern is the abstract shows no results and conclusions. A structured one might be better if possible.