Multimedia Digital Rights Protection Using Watermarking Techniques

Rao, N. Vyaghreswara; Pandit, Soumya

doi:10.1080/10658980701322528

Cited by 4 publications

(2 citation statements)

References 10 publications

(6 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, these well‐trained deep models with commercial value are vulnerable to malware infection or insider attacks, which may lead to the model being stolen and illegally used by malicious users. Given this model infringement problem, some researchers have recently proposed methods to prove a model copyright with digital watermarking, 2‐5 which embeds a watermark into the redundant information of the model. However, the model watermark cannot prevent the model from being used illegally in advance.…”

Section: Introductionmentioning

confidence: 99%

A framework for deep neural network multiuser authorization based on channel pruning

Wang

Song

Zhu

et al. 2023

Concurrency and Computation

View full text Add to dashboard Cite

SummaryVarious deep neural network (DNN) model watermarks have been proposed by researchers to verify copyrights for deep neural networks DNN. However, most DNN watermarking methods cannot prevent attackers from stealing and using the model. Unlike many existing approaches, this paper uses a channel pruning algorithm to protect DNN models, which verifies DNN models copyrights but also prevents the illegal use of DNN models. In this work, the pruning threshold or pruning rate is used as the secret key of a DNN model. After the secret key is distributed to multiple users, they prune the DNN model with the secret key, and the pruned and fine‐tuned model is provided to the users. The users can verify ownership of the model according to the pruning accuracy and fine‐tuning accuracy. If the secret key is incorrect, the accuracy of the model after fine‐tuning will be very low, and users will be unable to use the reasoning function of the fine‐tuned model. Based on the CIFAR‐10 and CIFAR‐100 datasets, we conducted experiments on five popular DNN models. The experimental results show that we can authorize multiple users by pruning very few channels in the convolution layers of the DNN model.

show abstract

Section: Introductionmentioning

confidence: 99%

A framework for deep neural network multiuser authorization based on channel pruning

Wang

Song

Zhu

et al. 2023

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…However, as digital products can be copied, stored and distributed without distortion, digital products can be reprinted on the Internet indiscriminately, resulting in many cases of piracy and infringement of the legitimate rights and interests of copyright owners of digital products. Therefore, how to protect the copyright of digital products is an urgent problem to be solved [32].…”

Section: Introductionmentioning

confidence: 99%

New watermarking algorithm utilizing quaternion Fourier transform with advanced scrambling and secure encryption

Bhatti

Yuan

et al. 2021

Multimed Tools Appl

View full text Add to dashboard Cite

Digital watermarking technology, as a powerful tool for copyright protection and content authentication of multimedia works, has received increasing attention, and the current image watermarking technology has developed significantly. Generally, embedding a watermark is done in grayscale images, mainly due to the fact that grayscale images are easier to process than color images, and grayscale images only contain brightness information and color-free information, in which an embedded watermark will not produce new color components. In real life, however, color images are more common than grayscale image information. To improve the security of the watermark information and the ability to embed the location and improve the security of the algorithm against a variety of attacks, this paper proposes an algorithm based on quaternion Fourier transform (QFT) with chaotic encryption and Arnold scrambling. Geometric Algebra (GA) can process color images in the form of vectors, and each component of Red,Green and Blue (RGB) should be processed separately. We utilized quaternion (which is a subalgebra of GA) and effectively completed color image processing by utilizing Fourier transform. After utilizing QFT, each component was made more secure by scrambling the pixels of the watermark and performing encryption utilizing chaotic sequencing. Different types of results utilizing Mean Square Error (MSE),Peak Signal to Noise Ratio (PSNR),Structure Similiarity Index Metric (SSIM), etc. were computed to verify the performance of our proposed algorithm and show that the robustness and security are high by utilizing our method.

show abstract

On Security of Compressed Gray Scale Image Using Joint Encryption and Data Hiding

Phadikar

Maity

2011

Information Security Journal: A Global Perspective

View full text Add to dashboard Cite

Multimedia Digital Rights Protection Using Watermarking Techniques

Cited by 4 publications

References 10 publications

A framework for deep neural network multiuser authorization based on channel pruning

A framework for deep neural network multiuser authorization based on channel pruning

New watermarking algorithm utilizing quaternion Fourier transform with advanced scrambling and secure encryption

On Security of Compressed Gray Scale Image Using Joint Encryption and Data Hiding

Contact Info

Product

Resources

About