Next Article in Journal
Agri-Ecological Policy, Human Capital and Agricultural Green Technology Progress
Next Article in Special Issue
Influence Relationship between Online News Articles and the Consumer Selling Price of Agricultural Products—Focusing on Onions
Previous Article in Journal
Self-Regulation of Soil Enzyme Activity and Stoichiometry under Nitrogen Addition and Plastic Film Mulching in the Loess Plateau Area, Northwest China
Previous Article in Special Issue
Design of 4UM-120D Electric Leafy Vegetable Harvester Cutter Height off the Ground Automatic Control System Based on Incremental PID
 
 
Article
Peer-Review Record

SSMDA: Self-Supervised Cherry Maturity Detection Algorithm Based on Multi-Feature Contrastive Learning

Agriculture 2023, 13(5), 939; https://doi.org/10.3390/agriculture13050939
by Rong-Li Gai *, Kai Wei and Peng-Fei Wang
Agriculture 2023, 13(5), 939; https://doi.org/10.3390/agriculture13050939
Submission received: 1 April 2023 / Revised: 14 April 2023 / Accepted: 19 April 2023 / Published: 25 April 2023

Round 1

Reviewer 1 Report

The figures need to be improved in general, for example titles of figure 7 including (a), (b), etc, should be referred in the caption, and also appropriately discussed inside the text about the colors and so on. Moreover, a short explanation about overview of the figure should be provided in caption, for example figure 2.

The main focus of Abstract should be on representative outcomes of the study with only a very short explanation of the importance of this study, as a motivation for the readers. Please, rewrite this section.

I read the Introduction section, and I could not convince, at least, myself that “Why are the self-supervised cherry maturity detection utilized?” Please, provide detailed information to answer this critical question.

Please, satisfy the readers with the most representative references in the Section 1 and 4 about the data.

Provide justifications on why some measures and algorithms are selected to detect the objects but not others.

Section 5: The results should be appropriately discussed with more information about the features that the approach was unable to detect during the simulations. It would be a valuable practice to provide useful suggestions to overcome these weaknesses.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper introduces a type of contrastive loss-based algorithm and object detection algorithm for a small-sized object database. The paper implements the self-supervised algorithm for cherry fruit detection with occlusion.

The sentence framing of the citation in Line 26 is not perfect.

Line No 148 has to be reframed.

The section 3 could have included a comprehensive explanation of the YOLO-v5 ST or other architectural elements.

The section 4 does not tabulate the values of parameters related to NT-Xent given in Equation 1 and 2.

In section 4 the experimentation value of Equation 7 is missing.

The ablation experimentation is clearly analyzed.

Minor Editing for Language correction is needed.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Back to TopTop