Emotion Recognition Using Convolutional Neural Network with Selected Statistical Photoplethysmogram Features
Round 1
Reviewer 1 Report
The paper is well written, easy to read and understand.
However, there are several issues to be solved:
- Section 2 (Emotion model) could be expanded to include explanations about Dominance/Control dimension (the authors only present Valence and Arousal dimensions in their study, but there are other studies in the literature where a third dimension is considered). Moreover, the authors only present categorical and dimensional emotion theories, but other theories could be found in the literature, such as the appraisal-based emotion theory.
- Please add a paragraph with more information about the computational cost. On table 5 the authors show the recognition intervals, but it is not clear whether this measure is reached with the same computing machine or not. Therefore, it is not clear whether techniques can be compared using that measure.
- Please make the units in each dimension of the figures known (specially Figures 3-6).
- Are the authors planning future works in this area? For example, including Dominance/Control dimension could be interesting to analyze. Please explain it.
Author Response
Revision Reports
Manuscript ID: applsci-795828
Title: Emotion recognition using convolutional neural network with selected statistical photoplethysmogram features
We would like to thank the editor and reviewers for their valuable comments and suggestions. The paper has been revised by considering the comments and we denoted the modified parts with red letters and line numbers of pdf. The detailed modifications in the revised version of the paper are in involved file.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The paper is interesting and deserves attention.
Before acceptance, I do have some suggestions to improve its quality, as acknowledged:
- Section 1:
- 1. The lines 23-27 are a bit confusing and need to be rephrased
- 2. Also, lines 35-37 include some repetitions. Please, reformulate
- 3. The authors mentioned the advantages of PPG with respect to ECG, SC, etc., but no drawbacks were cited. In order to write in a more unbiased manner, it is desirable to include also them
- 4. Some more recent literature about biosignals and emotions recognition is needed
- 5. Figure 1 should be placed where is cited in the text
- Section 3:
- 1. It is quite challenging to extract HRV from PPG: do you have any reference literature about this methodology to be cited?
- Section 4:
- 1. I think the discussion should have a separate paragraph and needs to be enhanced
- Section 5:
- 1. I think it is quite ambitious to recognize emotions in 10s windows. Please, discuss citing prior publications
- 2. A stronger take-home message and possible practical applications of this approach should be included
- Overall, grammar and language should be revised and typos should be corrected.
Author Response
Revision Reports
Manuscript ID: applsci-795828
Title: Emotion recognition using convolutional neural network with selected statistical photoplethysmogram features
We would like to thank the editor and reviewers for their valuable comments and suggestions. The paper has been revised by considering the comments and we denoted the modified parts with red letters and line numbers of pdf. The detailed modifications in the revised version of the paper are in involved file.
Author Response File: Author Response.docx
Reviewer 3 Report
An interesting article on recognizing emotions. The article is written clearly. Compared to the other works mentioned in the article, the progress is visible, although not so significant. Nevertheless, I would like the author to add some information:
- Why did you use only 240 of PPG samples (about 2s) and not 1280 samples(10s)? Is the other part of the signal irrelevant or not? (lines 199 - 201)
- Why were 80% of the samples used for learning and 20% for testing? Did the authors try another ratio of sample distribution? What was the result? (line 254)
- How many training cycles have been implemented? (line 256)
I recommend the authors to take a larger group of people because set of 20 people is not significant for drawing relevant conclusions.
Author Response
Revision Reports
Manuscript ID: applsci-795828
Title: Emotion recognition using convolutional neural network with selected statistical photoplethysmogram features
We would like to thank the editor and reviewers for their valuable comments and suggestions. The paper has been revised by considering the comments and we denoted the modified parts with red letters and line numbers of pdf. The detailed modifications in the revised version of the paper are in involved file.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors have considered previous comments given by the reviewers, therefore I suggest accepting this paper.
Reviewer 2 Report
All my concerns were successfully answered.
