Next Article in Journal
Editorial for Special Issue “Virtual Reality and Its Application in Cultural Heritage”
Next Article in Special Issue
A New Real-Time Detection and Tracking Method in Videos for Small Target Traffic Signs
Previous Article in Journal
Assessing the Technical Specifications of Predictive Maintenance: A Case Study of Centrifugal Compressor
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Applied Sciences
Volume 11
Issue 4
10.3390/app11041529
Review Report
applsci-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
Article
Peer-Review Record

Wheel Hub Defects Image Recognition Base on Zero-Shot Learning

Appl. Sci. 2021, 11(4), 1529; https://doi.org/10.3390/app11041529
by Xiaohong Sun 1,2, Jinan Gu 1,*, Meimei Wang 2,*, Yanhua Meng 2 and Huichao Shi 2
Reviewer 1: Anonymous
Reviewer 2:
Hantao Huang
Appl. Sci. 2021, 11(4), 1529; https://doi.org/10.3390/app11041529
Submission received: 4 January 2021 / Revised: 26 January 2021 / Accepted: 1 February 2021 / Published: 8 February 2021
(This article belongs to the Special Issue Artificial Intelligence for Computer Vision)

Round 1

Reviewer 1 Report

  1. Grammar mistakes and typos, for example:
    • In abstract, line 6: Therefore, in this paper, a generalized zero-shot learning framework for hub defect image recognition.
    • Page 2, 2nd paragraph, line 2: In order to identify unseen classes, we usually use a large labeled samples from
    • page 16, line 2: And in Figure 7 shows the effect of domain adaptation by using
    • page 16, 2nd paragraph: It can be seen from Table 3 that the performance improvement between DEM+RAC and DEM+RA is GU by 11.4
  2. The problem "Wheel Hub Defects recognition" is not explained in the introduction.
  3. The literature review is also very generic, it lacks the previous work done on the specific problem of "wheel hub defects recognition".
  4. What is the originality of this work?
  5. The author's contribution is not clear.
  6. In table 3. harmonic mean H is used to report the performance. It is not explained whether the accuracy used for computing H includes both seen and unseen classes or just unseen classes.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The major problem of this paper is writing. The motivation is not clear as well as the contribution. More details are:

 

1. If semantic descriptor exists, would it possible to use the semantic descriptor to determine the defect types ?

2. If you has some information of unseen defects and extract features from that information, will the be treated as zero shot learning ?

3. Why we need reverse mapping?

4. Writing format has problem at Section 4.2. E.g. corressp_o_nding

5. In alg2 box, it is three-step algorithm. However, you can not find any three steps in the bounding box. 

6. Reference 6 is missing key information. 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

I appreciate the changes authors have made in response to the raised concerns. 
I am satisfied with author's response and find it appropriate to be accepted to this journal.

Back to TopTop