Next Article in Journal
Evaluation of the Biomechanical Parameters of Human-Wheelchair Systems during Ramp Climbing with the Use of a Manual Wheelchair with Anti-Rollback Devices
Previous Article in Journal
Comparisons of the Anti-Inflammatory Activity of Dendropanax morbifera LEV Leaf Extract Contents Based on the Collection Season and Concentration of Ethanol as an Extraction Solvent
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Applied Sciences
Volume 10
Issue 23
10.3390/app10238754
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

Hybrid Local and Global Deep-Learning Architecture for Salient-Object Detection

Appl. Sci. 2020, 10(23), 8754; https://doi.org/10.3390/app10238754
by Wajeeha Sultan 1, Nadeem Anjum 1,*,†, Mark Stansfield 2 and Naeem Ramzan 2,†
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Appl. Sci. 2020, 10(23), 8754; https://doi.org/10.3390/app10238754
Submission received: 28 October 2020 / Revised: 26 November 2020 / Accepted: 27 November 2020 / Published: 7 December 2020
(This article belongs to the Section Computing and Artificial Intelligence)

Round 1

Reviewer 1 Report

The paper presents a deep learning approach with local and global features for salient object detection. The overall work is presented in a clear manner with background literature, explanations of the method, and sufficient comparisons. Minor comments:

- Please go through the text for minor typos.
- Some references are not numbered in the main text, e.g. line 106.

Author Response

Please see the attachment.

Author Response File: Author Response.doc

Reviewer 2 Report

  1. The English writing of this article needs to be revised carefully. There are many grammatical errors in this paper.
  2. One of the main contributions presented in this work is that the enhanced performance of salient object detection in low-contrast images. Please provide some experiments or examples to support this claim.
  3. Please check the correctness of eq. (1). The letter i seems to be the subscript of capital F. Besides, in this equation, how to calculate the local average F*?
  4. There are some strange operations in eqs. (2) and (3). For example, the function Upsample[], CAT(), and Conv(). Some of them can be achieved by layers of a network, and some of them need to be pre-defined by a mathematical operation. It is suggested to describe them in details.
  5. In lines 148-149, the sentence “three convolutional layers… and 512 dilations (where dilation =2) is applied”. Why 512 dilations are used here?
  6. In line 190, “The decoder follows same structure as encoder.” Is it correct? In general, the structure between them is symmetrical, not the same.
  7. Please improve the quality of figure 5. It is hard to observe.
  8. Please check table 2. The last two columns seem incorrect. Besides, why using the metric weighted F measure (WF) in this numerical comparison?

Generally, this paper proposes a deep-learning-based model to detect the salient object region pixelwisely. The results seems interesting, and worthy for publishing. However, the quality does not meet the standard of a periodical. Please improve this manuscript.

Author Response

Please see the attachement.

Author Response File: Author Response.doc

Round 2

Reviewer 2 Report

This version is much better than the previous one. Most parts of this paper are fine and seem having certain contribution of salient object detection. However, the quality of Figs. 6 and 7 are blurred. Please reproduce the figures. 

Author Response

Please see the attachment.

Author Response File: Author Response.doc

Back to TopTop