Lightweight Privacy Protection via Adversarial Sample
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis paper is well-structured and targets an interesting topic: adversarial sample based privacy protection. The authors proposed a method that performs locally, and therefore, by removing any involvement of a third party, provides privacy. Moreover, the method considers the local-device's limitations, by being lightweight. The manuscript is complete, and there are only minor revisions are required:
Line 52: Please explain this sentence in more details: ''Further clarification is needed here.''
Lines 119, 237, etc. : In academic writing, it is strongly advised to add a brief introduction in the beginning of each section.
Comments on the Quality of English LanguageThere are minor errors. For example, the Section headings are sometimes capitalized and sometimes not.
There are several places that the article 'the' is missing. For instance, the attack model.
There are some typos such as preotection.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThis paper attempted to reduce the requirements of adversarial sample privacy protections on devices, making the privacy protection more locally friendly.
-- "it has been demonstrated that adversarial sample-based 40 privacy protections achieve a better balance in terms of computational complexity, security, and data utility, making them more effective for privacy protection" where is this conclusion stated?
-- What is the challenge of "applying privacy protection based on adversarial samples locally to better serve its privacy protection function"?
-- What is the criteria for choosing the four datasets?
-- Could you please provide a Table to compare this one with existing works?
-- How the security is assessed for the given solution?
Author Response
Please see the attachment.
Author Response File: Author Response.doc