CIRO: The Effects of Visually Diminished Real Objects on Human Perception in Handheld Augmented Reality

Round 1

Reviewer 1 Report

The paper presents two contributions to the visual diminishment of real objects in handheld augmented reality. The first one is an algorithm to remove content-irrelevant real objects (CIRO) from the scene. The algorithm combines two already existing solutions (VINet and YOLACT) and is limited to certain types of human interference (as specified in Section 6.5). The second contribution is an experiment, where effects of the visual diminishment on human perception are studied. Three different situations are considered: a perfect diminishment (total removal of the CIRO),   no diminishment (CIRO remains in the scene) and an imperfect diminishment, using the algorithm. 

The study seems to be performed correctly and the results are interesting, albeit not surprising.  

However, there are some points that need modifications: 

1. On lines 273-274 you say that each participant of the experiment “was asked to count the number of jumps made by the virtual pet as a way to measure their concentration to the main AR content.” This assumes that the participant watched each video only once. Please, specify how did you ensure that the participants were not able to re-watch the videos. 

1. Please, specify how the individual options (points) in the used Likert scales were named. In addition, please, explain why you used the 5-point scale for the object realism question and the 7-point scale for other questions. 

1. Please, consider changing the shortcut DF for one of the test conditions. In my opinion, DF (no diminishment) is quite confusing as it is similar to DR (diminished reality). 

1. There are some minor language issues that should be corrected. The lines with ones detected by the reviewer are listed below (with the corresponding parts in brackets): 1 (scenes often), 6 (contents), 20 (are by nature dynamic), 21 (appear into the scene) 23 (innocent), 30 (applying), 51 (intrude), 55 (outlines), 106-107 (methodologies ... replaces), 116&121 (VINet VINET), 192 (makes), 259 (questionnaires 5.3), 260 (could the subjects proceed), 299 (it might be show up), 313 (just could bear), 311-314 (The sentence is too long.), 387 (simple assume movement), 405-406 (implementations...it should), 411 (to extend the continue our work). 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Please address some comments made on the uploaded file.

Comments for author File: Comments.pdf

Author Response

First of all, we thank the reviewers for their valuable comments. 
We accept most of the feedback and revise the paper as recommended. 
We give our response to the comments.

Comment 1 : (276 Line) How were they recruited, randomly?  

Response 1: 
We recruited participants from Korea University and Korea Institute of Science and Technology (KIST). 
We encouraged participation by using an internal e-mail service (mailing group composed of students) and SNS posts from the community where only those with affiliation accounts can sign up.
We collected the e-mail addresses of those who wanted to participate. 
When we received the e-mail, we explained the experiment to the subject and then provided a website link with experimental descriptions.
Thus our participants were randomly recruited, but the experimental process was controlled.

We hope that this is understandable for our study.

And Finally, we attached the revised paper you provided.
Please let us know if you have any questions or concerns.

Sincerely.

Author Response File: Author Response.pdf