2020
DOI: 10.1109/access.2020.3039323
|View full text |Cite
|
Sign up to set email alerts
|

DeepTrigger: A Watermarking Scheme of Deep Learning Models Based on Chaotic Automatic Data Annotation

Abstract: With the rapid development of artificial intelligence, the intellectual property protection of deep learning models appeals widespread concerns of scientists and engineers. The black-box watermarking protection scheme has been favored by many scholars due to its many advantages. The trigger set containing data content and data annotation is the key of black-box watermarking technology. However, most of the trigger sets in literates were constructed by comprehensible features, such as Gaussian noise and badges … Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
7
0
1

Year Published

2021
2021
2024
2024

Publication Types

Select...
7
1

Relationship

0
8

Authors

Journals

citations
Cited by 15 publications
(8 citation statements)
references
References 20 publications
0
7
0
1
Order By: Relevance
“…Unstructured data refer to data with variable length and no fixed format. Data expression and products of educational resources based on machine learning are usually protected by copyright in the early days, but copyright protects the expression form of ideas, the data expression and the selection, arrangement, system, and structure of the work, rather than the protection of the data itself [9]. And if the data are provided in the form of public dissemination, it will inevitably be exposed to and obtained by an unspecified Complexity number of people, and the service provider cannot guarantee that its application purpose must comply with the rules, which creates the risk of copyright infringement.…”
Section: Intellectual Property Expression Of Educational Resource Data Based On Machine Learningmentioning
confidence: 99%
“…Unstructured data refer to data with variable length and no fixed format. Data expression and products of educational resources based on machine learning are usually protected by copyright in the early days, but copyright protects the expression form of ideas, the data expression and the selection, arrangement, system, and structure of the work, rather than the protection of the data itself [9]. And if the data are provided in the form of public dissemination, it will inevitably be exposed to and obtained by an unspecified Complexity number of people, and the service provider cannot guarantee that its application purpose must comply with the rules, which creates the risk of copyright infringement.…”
Section: Intellectual Property Expression Of Educational Resource Data Based On Machine Learningmentioning
confidence: 99%
“…Например, существует схема встраивания, которая основана на присвоении каждому триггеру случайной метки [10,11]. В [12] набор триггеров состоит из абстрактных изображений.…”
Section: схемы встраивания Black-boxunclassified
“…Due to these features, there has been a rise in research attempting to embed digital chaos into cryptography. Various cryptographic applications have adopted digital chaos in their designs such as image encryption [1]- [6], S-box [7], [8], image watermarking [9], [10], pseudo-random number generators [11], [12], security protocols [13], [14], and hash function [15]- [17]. There are also other chaos-based applications that have been proposed such as secure speech encryption based on chaotic map [18], image encryption [19], and chaos-based compressive sensing encryption [20].…”
Section: Introductionmentioning
confidence: 99%