DeepKeyGen: A Deep Learning-Based Stream Cipher Generator for Medical Image Encryption and Decryption

Ding, Yi; Tan, Fengxiao; Qin, Zhen; Cao, Mingsheng; Choo, Kim‐Kwang Raymond

doi:10.1109/tnnls.2021.3062754

Cited by 79 publications

(40 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As shown in Table 6 which displays the comparative results assessing our proposed system with older systems due to their NPCR and UACI values, from the following analysis, it has been proved that the proposed scheme is robust and can withstand the differential assaults. [35] 0.9961 0.3357 Ding et al [23] 0.9959 0.2319 Bao and Xue [39] 0.9964 0.3349…”

Section: Resistance To Differential Attack Attackersmentioning

confidence: 99%

A deep learning signed medical image based on cryptographic techniques

Elkamchouchi

Algarni

Ghoniem

et al. 2022

IJEECS

View full text Add to dashboard Cite

Innovative medical multimedia communications technology requirements have enhanced safety principles, allowing significant advancement in security standards. In hospitals and imaging centers, massive amounts of medical images have been created. To successfully access the medical databases and utilize those rich resources in assisting diagnosis and research, image processing enabled communication solutions are necessary. Our article presents a rigorous verified model by employing deep learning to enhance the cryptographic performance of biomedical images using hybrid chaotic Lorentz map diffusion and de-oxyribonucleic acid (DNA) confusion stages. It consists of two encryption/decryption techniques, the initial signal is verified using digital signature and two unique non-consecutive stages of chaotic diffusion with a single DNA scrambling stage in between. The encoded secret bit stream is generated and used to encrypt or decode the original signal in the diffusion manner to disintegrate the redundancy in the plain image statistics, utilizing hybrid chaotic system. Using DNA confusion step to make the relationship between the original signal and the utilized key more ambiguous. These stages make the proposed image cryptosystem more resistant to known/ chosen plaintext assaults. The performance of the suggested technique will be assessed to the most similar techniques reported in the literature for comparative purposes.

show abstract

Section: Resistance To Differential Attack Attackersmentioning

confidence: 99%

A deep learning signed medical image based on cryptographic techniques

Elkamchouchi

Algarni

Ghoniem

et al. 2022

IJEECS

View full text Add to dashboard Cite

show abstract

“…where κ is a parameter of chaotic map, which is used to adjust the value of β-map and indicate the bifurcation parameter. Beta function is utilized in various types of applications in image processing and engineering, especially in bio-medical signal and image compression, image detection [27][28][29][30][31].…”

Section: Fractional Beta Chaotic (Fbc) Mapsmentioning

confidence: 99%

“…Finally, Arnold map and RSA algorithm are used by Zahmoul et al [26]. Beta function is utilized in various types of applications in image processing and engineering, especially in bio-medical signal and image compression, image detection [27][28][29][30][31][32][33][34][35][36]. Other works can be located in [37][38][39][40][41][42][43][44][45][46][47][48][49].…”

Section: Introductionmentioning

confidence: 99%

Image Encryption Algorithm Based on New Fractional Beta Chaotic Maps

Ibrahim¹,

Natiq²,

Alkhayyat³

et al. 2022

Computer Modeling in Engineering &Amp; Sciences

View full text Add to dashboard Cite

In this study, a new algorithm of fractional beta chaotic maps is proposed to generate chaotic sequences for image encryption. The proposed technique generates multi random sequences by shuffling the image pixel position. This technique is used to blur the pixels connecting the input and encrypted images and to increase the attack resistance. The proposed algorithm makes the encryption process sophisticated by using fractional chaotic maps, which hold the properties of pseudo-randomness. The fractional beta sequences are utilized to alter the image pixels to decryption attacks. The experimental results proved that the proposed image encryption algorithm successfully encrypted and decrypted the images with the same keys. The output findings indicate that our proposed algorithm has good entropy and low correlation coefficients. This translates to enhanced security against different attacks. A MATLAB programming tool was used to implement and assess the image quality measures. A comparison with other image encryption techniques regarding the visual inspection and signal-to-noise ratio is provided.

show abstract

“…D EEP learning methods have been widely adopted in a broad range of domains, such as image classification [1][2][3][4], image security [5][6][7], medical aided diagnosis [8][9][10][11] and facial recognition [12,13]. In some of these application domains, the performance of deep learning models has reportedly exceed those of human.…”

Section: Introductionmentioning

confidence: 99%

Interpreting Universal Adversarial Example Attacks on Image Classification Models

Ding

Tan

Geng

et al. 2023

IEEE Trans. Dependable and Secure Comput.

Self Cite

View full text Add to dashboard Cite

Mitigating adversarial deep learning attacks remains challenging, partly because of the ease and low cost in carrying out such attacks. Therefore, in this paper, we focus on the understanding of universal adversarial example attack on image classification models. Specifically, we seek to understand the difference(s) between adversarial examples in two adversarial datasets (DAmageNet and PGD dataset) and clean examples in ImageNet learned by the classification model, and whether we can use such findings to resist adversarial example attacks. We also seek to determine if we can retrain a discriminator to discriminate whether the input image is an adversarial example, using adversarial training. We then design a number of experiments (e.g., class activation map (CAM) analysis, feature map analysis, feature maps/filters changing, adversarial training, and binary classification model) to help us determine whether the universal adversarial dataset can be successfully used to attack the classification model. This, in turn, contributes to a better understanding of adversarial defenses over pretrained classification model from an interpretation perspective. To the best of our knowledge, this work is one of the earliest works to systematically investigate the interpretation of universal adversarial example attack on image classification models, both visually and quantitatively.

show abstract

DeepKeyGen: A Deep Learning-Based Stream Cipher Generator for Medical Image Encryption and Decryption

Cited by 79 publications

References 33 publications

A deep learning signed medical image based on cryptographic techniques

A deep learning signed medical image based on cryptographic techniques

Image Encryption Algorithm Based on New Fractional Beta Chaotic Maps

Interpreting Universal Adversarial Example Attacks on Image Classification Models

Contact Info

Product

Resources

About