اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدDeepiSign: Invisible Fragile Watermark to Protect the Integrityand Authenticity of CNN
79
0
0.0
(
0
)
نشر من قبل
Alsharif Abuadbba Dr
تاريخ النشر
2021
مجال البحث
الهندسة المعلوماتية
والبحث باللغة
English
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Convolutional Neural Networks (CNNs) deployed in real-life applications such as autonomous vehicles have shown to be vulnerable to manipulation attacks, such as poisoning attacks and fine-tuning. Hence, it is essential to ensure the integrity and authenticity of CNNs because compromised models can produce incorrect outputs and behave maliciously. In this paper, we propose a self-contained tamper-proofing method, called DeepiSign, to ensure the integrity and authenticity of CNN models against such manipulation attacks. DeepiSign applies the idea of fragile invisible watermarking to securely embed a secret and its hash value into a CNN model. To verify the integrity and authenticity of the model, we retrieve the secret from the model, compute the hash value of the secret, and compare it with the embedded hash value. To minimize the effects of the embedded secret on the CNN model, we use a wavelet-based technique to transform weights into the frequency domain and embed the secret into less significant coefficients. Our theoretical analysis shows that DeepiSign can hide up to 1KB secret in each layer with minimal loss of the models accuracy. To evaluate the security and performance of DeepiSign, we performed experiments on four pre-trained models (ResNet18, VGG16, AlexNet, and MobileNet) using three datasets (MNIST, CIFAR-10, and Imagenet) against three types of manipulation attacks (targeted input poisoning, output poisoning, and fine-tuning). The results demonstrate that DeepiSign is verifiable without degrading the classification accuracy, and robust against representative CNN manipulation attacks.
قيم البحث
اقرأ أيضاً
Deep learning techniques have made tremendous progress in a variety of challenging tasks, such as image recognition and machine translation, during the past decade. Training deep neural networks is computationally expensive and requires both human an
d intellectual resources. Therefore, it is necessary to protect the intellectual property of the model and externally verify the ownership of the model. However, previous studies either fail to defend against the evasion attack or have not explicitly dealt with fraudulent claims of ownership by adversaries. Furthermore, they can not establish a clear association between the model and the creators identity. To fill these gaps, in this paper, we propose a novel intellectual property protection (IPP) framework based on blind-watermark for watermarking deep neural networks that meet the requirements of security and feasibility. Our framework accepts ordinary samples and the exclusive logo as inputs, outputting newly generated samples as watermarks, which are almost indistinguishable from the origin, and infuses these watermarks into DNN models by assigning specific labels, leaving the backdoor as the basis for our copyright claim. We evaluated our IPP framework on two benchmark datasets and 15 popular deep learning models. The results show that our framework successfully verifies the ownership of all the models without a noticeable impact on their primary task. Most importantly, we are the first to successfully design and implement a blind-watermark based framework, which can achieve state-of-art performances on undetectability against evasion attack and unforgeability against fraudulent claims of ownership. Further, our framework shows remarkable robustness and establishes a clear association between the model and the authors identity.
Recently, a self-embedding fragile watermark scheme based on reference-bits interleaving and adaptive selection of embedding mode was proposed. Reference bits are derived from the scrambled MSB bits of a cover image, and then are combined with authen
tication bits to form the watermark bits for LSB embedding. We find this algorithm has a feature of block independence of embedding watermark such that it is vulnerable to a collage attack. In addition, because the generation of authentication bits via hash function operations is not related to secret keys, we analyze this algorithm by a multiple stego-image attack. We find that the cost of obtaining all the permutation relations of $lcdot b^2$ watermark bits of each block (i.e., equivalent permutation keys) is about $(lcdot b^2)!$ for the embedding mode $(m, l)$, where $m$ MSB layers of a cover image are used for generating reference bits and $l$ LSB layers for embedding watermark, and $btimes b$ is the size of image block. The simulation results and the statistical results demonstrate our analysis is effective.
Recent research has demonstrated that adding some imperceptible perturbations to original images can fool deep learning models. However, the current adversarial perturbations are usually shown in the form of noises, and thus have no practical meaning
. Image watermark is a technique widely used for copyright protection. We can regard image watermark as a king of meaningful noises and adding it to the original image will not affect peoples understanding of the image content, and will not arouse peoples suspicion. Therefore, it will be interesting to generate adversarial examples using watermarks. In this paper, we propose a novel watermark perturbation for adversarial examples (Adv-watermark) which combines image watermarking techniques and adversarial example algorithms. Adding a meaningful watermark to the clean images can attack the DNN models. Specifically, we propose a novel optimization algorithm, which is called Basin Hopping Evolution (BHE), to generate adversarial watermarks in the black-box attack mode. Thanks to the BHE, Adv-watermark only requires a few queries from the threat models to finish the attacks. A series of experiments conducted on ImageNet and CASIA-WebFace datasets show that the proposed method can efficiently generate adversarial examples, and outperforms the state-of-the-art attack methods. Moreover, Adv-watermark is more robust against image transformation defense methods.
In the Internet-of-Things, the number of connected devices is expected to be extremely huge, i.e., more than a couple of ten billion. It is however well-known that the security for the Internet-of-Things is still open problem. In particular, it is di
fficult to certify the identification of connected devices and to prevent the illegal spoofing. It is because the conventional security technologies have advanced for mainly protecting logical network and not for physical network like the Internet-of-Things. In order to protect the Internet-of-Things with advanced security technologies, we propose a new concept (datachain layer) which is a well-designed combination of physical chip identification and blockchain. With a proposed solution of the physical chip identification, the physical addresses of connected devices are uniquely connected to the logical addresses to be protected by blockchain.
Training deep neural networks from scratch could be computationally expensive and requires a lot of training data. Recent work has explored different watermarking techniques to protect the pre-trained deep neural networks from potential copyright inf
ringements. However, these techniques could be vulnerable to watermark removal attacks. In this work, we propose REFIT, a unified watermark removal framework based on fine-tuning, which does not rely on the knowledge of the watermarks, and is effective against a wide range of watermarking schemes. In particular, we conduct a comprehensive study of a realistic attack scenario where the adversary has limited training data, which has not been emphasized in prior work on attacks against watermarking schemes. To effectively remove the watermarks without compromising the model functionality under this weak threat model, we propose two techniques that are incorporated into our fine-tuning framework: (1) an adaption of the elastic weight consolidation (EWC) algorithm, which is originally proposed for mitigating the catastrophic forgetting phenomenon; and (2) unlabeled data augmentation (AU), where we leverage auxiliary unlabeled data from other sources. Our extensive evaluation shows the effectiveness of REFIT against diverse watermark embedding schemes. In particular, both EWC and AU significantly decrease the amount of labeled training data needed for effective watermark removal, and the unlabeled data samples used for AU do not necessarily need to be drawn from the same distribution as the benign data for model evaluation. The experimental results demonstrate that our fine-tuning based watermark removal attacks could pose real threats to the copyright of pre-trained models, and thus highlight the importance of further investigating the watermarking problem and proposing more robust watermark embedding schemes against the attacks.
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
