Review History


All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.

Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.

View examples of open peer review.

Summary

  • The initial submission of this article was received on July 3rd, 2021 and was peer-reviewed by 3 reviewers and the Academic Editor.
  • The Academic Editor made their initial decision on July 29th, 2021.
  • The first revision was submitted on September 4th, 2021 and was reviewed by 1 reviewer and the Academic Editor.
  • The article was Accepted by the Academic Editor on November 11th, 2021.

Version 0.2 (accepted)

· Nov 11, 2021 · Academic Editor

Accept

You have addressed all the comments of the reviewers, and upon my reading of the paper it is ready for publication.

Reviewer 3 ·

Basic reporting

The authors have enhanced the article with more introductions and discussions.

Experimental design

No comment.

Validity of the findings

No comment.

Version 0.1 (original submission)

· Jul 29, 2021 · Academic Editor

Minor Revisions

Three reviewers think this paper needs some minor revisions.

[# PeerJ Staff Note: Please ensure that all review and editorial comments are addressed in a response letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.  It is a common mistake to address reviewer questions in the response letter but not in the revised manuscript. If a reviewer raised a question then your readers will probably have the same question so you should ensure that the manuscript can stand alone without the response letter.  Directions on how to prepare a response letter can be found at: https://peerj.com/benefits/academic-rebuttal-letters/ #]

Reviewer 1 ·

Basic reporting

The writing of the paper manuscript is clear and well organized. But there are some problems to be addressed. The citation of some references should be complete, including pages (e.g. Srinivasan, R., & Martinez, A. M. (2018).) Some of the figures should be resized (e.g. Figure 6).

Experimental design

The implementation of the methods should be described with sufficient detail for replication, including the setting of parameters. More recently state-of-the-art methods on supervised and unsupervised learning should be considered for comparison.

Validity of the findings

The experimental results are provided. But the contribution and novelty of the article are not clear.

Additional comments

The authors investigated the supervised and unsupervised emotion classifier performance. The comments should be addressed before further processing.

Reviewer 2 ·

Basic reporting

The content of this article is acceptable on the general with novel research topic and abundant detailed experimental results. Yet it still requires certain revisions to meet the publication standard, which are listed below in several aspects.
1. Language description: The descriptions in this article are unclear and hard to unsderstand at multiple places, such as Lines 56-57, 105-110, 229-231, 237-239, 372-374, etc. Please check the language used in the article.
2. Literature references: This article provided various prior works and references, however, the methods compared in the experiments are not fully included. For example, the late fusion methods mentioned (maximum rule, sum rule, product rule, weight criterion, rule-based, and model-based methods) have not been referenced, and the early fusion methods have not been explicited presented. Please properly cite all evaluated works in this paper.
3. Article structure: This article is organized in clear and reasonable structure. However, the discussion section should be more compact and conclusive, highlighting the major findings and contributions of this article.

Experimental design

Various experiments have been conducted for this research. Minor improvements are suggested as follows:
1. The optimal numbers of clusters for the traditional unsupervised approaches are said to be 2 or 6 through CH index and Silhouette score analysis (Lines 220-225). Experimental results should be presented to validate this statement.
2. In feature extraction and pre-processing section, certain data cleaning process has been conducted (Line 148). Please explain the reason for such process.
3. Multiple features for AUs are computed as video features (Lines 154-157), please give explicit definition or computational detail of each feature.

Validity of the findings

no comment

Reviewer 3 ·

Basic reporting

Very good paper. A few slight comments:
I understand that one of the strength of this study is to investigate the variety of emotional categories and to show the importance of each features (voice + face).

A little more introduction should be discussed in comparison with previous studies (see https://github.com/EvelynFan/AWESOME-MER) that adopted a multimodal approach.

Knowledge of features goes to application to robots. I felt that it would be good if there was a discussion about Human-robot interaction in the Discussion section.

Barrett et al (2019: Line 80-81) strictly criticizes emotion recognition in emotional categories like this paper. Citing Keltner is preferable as a position, so I recommend to cite it (e.g., https://psycnet.apa.org/record/2017-30838-004; http://emotionresearcher.com/the-great-expressions-debate/)

Experimental design

no comment

Validity of the findings

"actor and gender specific aspects also contributed to clustering"
What are the implications of this finding?
The Unsupervised method seems just to report the characteristics of the stimulus (GEMEP). I want the author to clarify that necessity.

All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.